Causes, consequences, and where we go from here
February 8, 2015
Shortly after midnight on August 4, 2014, the tailings storage facility (TSF) at the Mount Polley copper and gold mine breached, releasing an estimated 17 million cubic metres of water and eight million cubic metres of fine sand and mine tailings into Polley Lake. From Polley Lake, this effluent flowed via Hazeltine Creek into Quesnel Lake, destroying trees and eroding landforms in its path and creating a cloud of sediment in the water column of the lake, necessitating a drinking water ban for several days in the neighbouring town of Likely and its environs.
With the release January 30 of the first of several independent reports into the disaster, we now know the cause of the breach was failure in the design of the dam embankment. But how did that failure go undiscovered for so long? How did two engineering companies, each of which served as Engineer of Record at some time during the design and/or construction of the dam embankment, both uncritically accept unwarranted assumptions about the layers of soil the dam was built upon? And more importantly, what has been the environmental cost of the breach? Have the resident fish populations been affected? Are they safe to eat? Were toxins released to the environment? Is the water safe to drink? What is the provincial government doing to prevent this from happening at one of the dozens of other BC mines employing TSFs similar to Mount Polley’s?
A design doomed to fail
The design of the tailings storage facility at Mount Polley is a common one, in use at many mines throughout the world. The facility is an earthen dam set into a hillside to create a container to hold tailings, the sand-sized particles of rock left over from mining. It also holds water left over from mining operations (which are very water intensive). The water serves two purposes: (1) it keeps air from interacting with the tailings (oxidized tailings can produce acid, which mobilizes metals in solution, which can leach into the environment); and (2) it keeps tailings from escaping the enclosure by preventing them from being carried away as dust on the wind.
But the TSF is not designed to be a water dam. Its primary purpose is to contain tailings, and the design of the TSF assumes that tailings make up a “beach” that completely surrounds the water and prevents it from coming into direct contact with the embankment. The independent investigation into the cause of the Mount Polley breach, conducted by engineers Dr. Norbert R. Morgenstern, Mr. Steven G. Vick, and Dr. Dirk Van Zyl, revealed that Mount Polley had a recurring problem maintaining the required beach along the entire perimeter of the embankment.
The mine was storing too much water in the dam relative to the amount of tailings, and this water kept eroding the beach and coming into contact with the dam wall. There was evidence of resultant cracks and internal erosion of the embankment, and yet the water management problem persisted for years. As much as this may raise questions about the management of the TSF on the part of Mount Polley Mine Corp., the investigating panel determined that this ultimately was not the cause of the disaster.
What was the cause was geology – or rather, a lack of understanding of the complexity of the subsurface geology on the dam site. The panel determined that the dam breach was caused when a section of the perimeter embankment slid and rotated along a layer of clay eight metres below the surface. Simply put, a landslide caused the embankment wall to drop 3.3 m. Because the level of water in the dam was only 2.3 m below the top of the embankment, the slide caused the water to begin overtopping the embankment, and the erosive power of that water quickly took out a section of the embankment approximately 100 m long.
The Upper Glaciolacustrine Unit (the layer of clay that ultimately gave way under the weight of the embankment wall) varies in thickness throughout the dam site: in some places, it is completely absent; in others, it can be up to 2.5 m thick. One of the thickest areas is precisely where the landslide occurred. While Knight-Piésold, the engineering firm in charge of designing and building the TSF from its inception in 1995 until 2011, drilled holes to investigate the soil where the dam was to be built, only four of the holes were deep enough to have hit the Upper Glaciolacustrine Unit. Those four holes were not in the area where the dam breached.
Knight-Piésold knew about the clay layer, but did not know they were building the embankment wall directly over the thickest part of this layer. More critically, they didn’t take into account how the strain on that layer would change over time, particularly with the additional weight of the ever-growing dam and its contents. Their calculations depended on that layer remaining compacted and drained, as it appeared in their initial investigations. However, when the clay was subjected to the pressure of all the extra rock and water on top of it, significant pore pressures (from the water content of the clay) built up, causing the clay to lose a significant amount of its strength. The shearing force became too much and the soil layer gave way.
Loading the gun and pulling the trigger
The panel’s job was not to assign blame for the disaster, but rather to uncover its cause. The facts the panel uncovered, however, tell a story of limited foresight and poor planning decisions on the part of both the mine operator and the Engineer of Record (Knight-Piésold from 1995-2011 and AMEC from 2011 until the breach). “There was little in the way of long-term planning or execution. This was most clearly displayed by the absence of an adequate water balance or water treatment strategy, and the overtopping failure that nearly resulted,” the panel commented, referring to an overtopping incident in May 2014 that required emergency action to decrease the level of water in the TSF. “Moreover, the related absence of a well-developed tailings beach violated the fundamental premise of the design as a tailings dam, not a water-storage dam.”
One of the most damning passages follows: “The same problem was apparent in production and scheduling for mine waste used in dam construction. The design was caught between the rising water and Mine plan, between the imperative of raising the dam and the scarcity of materials for building it. Something had to give, and the result was oversteepened dam slopes, deferred buttressing, and the seemingly ad hoc nature of dam expansion that so often ended up constructing something different from what had originally been designed.”
In Stage 4 of dam construction (2006-2007), Knight-Piésold proposed a design for the perimeter embankment that incorporated a slope ratio of 2 horizontal units for every 1 vertical unit. Given what is now known about the undrained strength of the critical clay layer, even with this slope angle, the factor of safety for this design was far below accepted engineering standards. Yet the 2:1 slope was never constructed. The embankment as built had a 1.3:1 ratio.
This much steeper slope was adopted as a “temporary” measure during Stage 5 construction and, for reasons that are unclear, became the permanent slope angle through Stage 9, which is the stage at which the breach occurred. The panel remarked that “if constructing unknowingly on the Upper GLU [glaciolacustrine unit] and not recognizing the potential undrained failure constituted loading the gun, building with a 1.3H:1V angle of repose slope over this stratum pulled the trigger.”
The environmental effects
Thankfully, the tailings released at Mount Polley were not highly reactive. Despite the high metallic content in the tailings themselves, those metals do not easily become mobilized in solution. Thus, if you can filter out the particulates (or wait for them to settle to the bottom of the water column, if they are in a standing body of water), the remaining water would be safe to drink.
Immediately after the tailings dam breach, residents of the surrounding area were told not to use tap water for either bathing or drinking, because the water contained tailings particulates. On August 5, the Ministry of Environment (MOE) sampled water from Quesnel Lake and the Quesnel River. The tests indicated that water samples from both sites met provincial and federal safe drinking water guidelines. Water samples from Polley Lake taken on August 7 were comparable to samples that predated the disaster.
A sample taken from the middle of the sediment plume in Quesnel Lake on August 7 – what the MOE termed “a worst case scenario” for drinking water, given that residents were advised not to drink cloudy or turbid water – showed slightly elevated phosphorous and aluminum levels compared with Health Canada guidelines. But samples taken from outside the plume indicated that the water was within safe drinking water guidelines. The Ministry has continued to test the water weekly since late September and these tests have consistently shown that the water is safe to consume.
As for aquatic life, the picture is a little less clear. Tests conducted immediately after the disaster showed little impact on aquatic life. For example, water tests on August 5 determined that levels of zinc exceeded chronic aquatic life guidelines in 2 of 5 samples from Quesnel River, but that levels of all other metals were below accepted guidelines. Samples taken August 6 in the same areas showed no contaminant concentrations in excess of aquatic life guidelines.
But weekly water testing has continued on the Quesnel River at Likely Bridge and at the Gravelle Ferry Bridge (downstream from Likely Bridge), and tests conducted as recently as December have shown “exceedances of acute guidelines for total copper, dissolved aluminum, chromium, and total phosphorous,” according to the MOE, indicating that “impacts to wildlife may be occurring.” These levels exceed those measured in the two months following the breach, because in the fall, Quesnel Lake began to “turn over,” meaning that its warmer, less turbid surface water began to cool and sink, mixing with the colder, more turbid water at depth. The content of that turbidity is the fine sediment from the tailings sand that washed from the dam to the lake during the breach. Now that the lake is more uniformly turbid, samples taken from the river (which are always taken at a depth of 0.5 m) will reflect the metallic content of the tailings.
However, it should be noted that the dissolved concentrations of these metals (except for aluminum) is much lower than the total concentrations. The higher total concentrations are associated mainly with the particulates, and “may not be as bioavailable as dissolved metals, thus decreasing the risk to aquatic life,” according to the MOE report on the findings.
BC First Nations groups were concerned enough to conduct their own investigation of fish populations. The First Nations Health Authority (FNHA) collected 45 fish samples – 35 harvested after the breach and 10 harvested pre-breach – and analyzed them for total metals. Results were then reviewed by toxicologists working on the assumption that the typical BC First Nations adult consumes a lifetime average of one bowl of salmon per day, and the typical toddler one-half bowl per day. In all cases, metals were below levels of human health risk.
But knowing the fish are safe to eat doesn’t alleviate all concerns regarding the impact of the metallic content of the tailings. Scientific studies have shown that copper can damage the olfactory organ in fish, leaving them unable to find their way to their birth streams at spawning time. Imperial Metals, the owner/operator of Mount Polley mine, hired Golder Associates, a geotechnical firm, to evaluate water samples from the impact zone for their potential to affect fish in this manner. Golder found that dissolved copper levels, even at their highest concentration in a sample location near Hazeltine Creek, were below the thresholds at which they would begin to impair olfactory function. Moreover, salmon did indeed return to their spawning areas in the Quesnel system later in the year.
The bulk of Imperial Metals’ efforts in the past six months have been directed toward monitoring water quality, mitigating erosion, and controlling the transport of tailings sediments. A silt curtain has been installed at the mouth of Hazeltine Creek to catch transported sediments before they reach Quesnel Lake, and water monitoring stations in Hazeltine Creek and the Quesnel River take weekly samples. Dataloggers have been deployed at two of the water monitoring stations, and these instruments take field measurements – of turbidity, pH, specific conductance, dissolved oxygen, and temperature – every 15 minutes. The company submits weekly reports on its remediation activities to the Ministry of Environment, and these reports are publically available via Imperial Metals’ website.
For now, Mount Polley mine remains on care and maintenance indefinitely.
Looking to the future
There are currently 98 TSFs at 60 metal or coal mines in BC – 31 of these at mines in operation or under construction and the remaining 67 at mines that are permanently closed or under care and maintenance. On August 18, the Chief Inspector of Mines ordered that each mine’s annual Dam Safety Inspection be reviewed by an independent, third-party professional engineer from a firm unassociated with the TSF. The third-party reviewers were required to comment on each dam’s overall condition and its consequence classification. For dams with consequence classifications of high, very high, or extreme, companies were required to update and test their Emergency Preparedness and Response Plans and report on the results of that test and lessons learned. The third-party reviews did not reveal any immediate safety concerns.
According to the independent investigative panel, however, the answer to zero dam breaches in the future lies not with increased governmental regulation. The panel concluded that increased regulatory intervention from the Ministry of Energy and Mines could not have prevented the Mount Polley disaster. “By definition, no amount of inspection can discover a hidden flaw,” they said. Government regulators, the panel cautioned, “cannot usurp the role of the designer.” Rather, the regulators rely on the expertise of the engineering firm and its assurances that a proper site investigation has been completed.
The way to get to zero dam failures, according to the panel, is by eliminating tailings dams altogether. “This can be achieved by storing the majority of the tailings below ground – in mined-out pits for surface mining operations or as backfill for underground mines,” the panel said. This would require integrating tailings planning into mine planning, which has not been common practice in the industry.
Short of storing tailings underground, the panel recommended technology to create filtered tailings – tailings that have been dewatered before they leave the mill. The dewatered tailings are then transported to the tailings facility, compacted, and blanketed with a dry cover material instead of water. This technology has been in use mainly in dry climates where water conservation is a high priority, but it has also been used at the Greens Creek mine in Alaska since start-up. The only reason it is not in general use is that it costs more than conventional tailings dam storage. But according to the panel’s recommendations, “safety attributes should be evaluated separately from economic considerations, and cost should not be the determining factor.”
The government response to the report
After the panel’s report was released, the provincial government responded by requiring all operating mines with TSF dams to provide a letter by June 30, 2015 confirming whether foundation soils similar to those found at Mount Polley exist below their TSFs. If they do, companies are required to show that their site investigations were sufficient to map out and understand both the extent and strength of those subsurface layers and that their design calculations properly accounted for those conditions. The Province also initiated a code review to determine how best to implement the panel’s recommendations.
The panel’s report is just the first of several investigations of the Mount Polley breach. The Chief Inspector of Mines is conducting an investigation into whether there were any violations of the Mines Act or the Health, Safety and Reclamation Code for Mines in BC. A third investigation, by the Conservation Officer Service, will determine whether any environmental or natural resource laws were broken, such as the Environmental Management Act or the Fisheries Act. These reports are due later this year.