abstract background

3rd International 21st Annual Minerals Council of Australia Environmental Workshop

160

INCORPORATION OF LANDFORM DESIGN AND EROSIQN MODELLING IN MINE PLANNING AT THE RANGER MINE, JABIRU NT Corinne Unger Superintendent, Research and Rehabilitation, ERA Environmental Services Pty Ltd, Darwin Ken Evans & Michael Saynor Environmental Research Institute of the Supervising Scientist, Jabiru Garry Willgoose The University of Newcastle

ABSTRACT

I;

Research at Ranger is directed toward long term stability of the final landform which will incorporate uranium mill tailings, waste rock and sub-economic ores. Mining of Pit #1 concluded during 1994 and mining of Orebody #3 is planned to commence in 1997. Mine planning for the development of Orebody #3 has required a revision of the final landform design. The landform design assists with the delineation of upper limits for stockpiles to avoid double handling and also defines the lateral limits of disturbance of the surrounding environment. The detailed drainage and slope design is the subject of modelling to assist with the refinement of this landform design and the assessment of alternatives. Long term erosional stability of the Ranger Mine post-mining landform has been modelled using the landform evolution model SIDERIA, based on parameters derived for unvegetated areas of waste rock. This has provided an indication of the likely changes in landform morphology due to erosion (ie. gully development etc) in the absence of vegetation and on compacted dump surfaces. During the 1994/95 wet season, data were gathered from revegetated areas of the Ranger waste dump on infiltration, runoff and suspended sediment under a range of rainfall events, to provide new input parameters to SIDERIA. These parameters reflect the changes in 'minesoil' erodibility and erosivity of rainfall due to the presence of vegetation. The use of these new parameters has quantified the reduced rates of erosion and the magnitude of the resultant gullieslvalleys which are likely to form on the final landform. The subsequent stage to this research was the design of four alternative final landform designs with assistance from existing mine planning software at Ranger Mine (DATAMINE), which provided the necessary digital terrain models (DTM's). These can be imported readily into SIDERIA for evaluation of the erosional changes over time. The design which performs best can then be modified via an iterative process to develop landform design criteria critical to the stability of the landform and these criteria will be built into the landform via progressive rehabilitation of the stockpiles.

BACKGROUND Research at Ranger is directed toward long term stability of the final landform which will incorporate uranium mill tailings, waste rock and sub-economic ores. Mining of Pit #1 concluded in 1994 and mining of Orebody #3 is planned to commence in 1997. Mine planning for the development of Orebody #3 required a revision of the final landform design. This landform design assists with the delineation of upper limits for stockpiles to avoid double handling and also defines the lateral limits of disturbance of the surrounding environment. The detailed drainage and slope design is the subject of modelling to assist with the refinement of this landform design and the assessment ef alternatives. Figure 1 shows the current mine layout. It is proposed that the waste dumps (indicated as northern and southern stockpiles) will extend into the catchment of Retention Pond 1 (north of the tailings dam) when mining commences in Orebody #3 (ERA: 1996). Previous erosion studies Since around 1986, erosion studies have been undertaken at the Ranger mine. Table 1 aims to summarise how this research has developed under the programs of research carried out by ERA,


161

Ranger Mine and ERISS (Environmental Research Institute of the Supervising Scientist previously ARRRI - Alligator Rivers Region Research Institute). Landform stability and erosion studies are an integral component of both research programs (Appendices 1 and 2). Initially, erosion studies were carried out on slopes of the northern waste dump. Manual sampling during the first trial revealed several limitations. Subsequent trials used autosamplers and flumes to confine runoff. Comparative slope trials were carried out over the next three wet seasons, though the rainfall conditions and associated erosion magnitude only reflected average and below-average conditions and there was a need to gain an estimation of conditions over the long term ie. 1 000 years. This time frame was important because of the requirement to rehabilitate uranium mill tailings in impoundments which have a design life of 200 years and a structural life of 1 000 years (AGPS: 1982). It was recognised that modelling of erosion had a role in predicting the magnitude of erosion over these time frames. The landform evolution model SIBERIA was then developed for the Ranger site. SIBERIA (Willgoose et al: 1989) is a computer model developed to study the erosional development of catchments and their channel/gully networks. Elevation changes are simulated by a mass transport continuity equation applied over geologic time. Mass transport includes fluvial sediment transport, such as modelled by the Einstein-Brown equation, and mass movement mechanisms such as creep, rainsplash or landslide. If more material enters a region than leaves it, then the elevations rise, and vice versa. The runoff and sediment transport data collected by ERISS for natural rainfall runoff events and rainfall simulator experiments were used for the calibration of the runoff and sediment transport models within SIBERIA, and SIBERIA used to simulate elevation changes over 1000 years. Willgoose and Riley (1993) found, with SIBERIA, the main difference with calibration of a traditional agricultural sediment transport model, is that the dependency of transport on discharge and slope need to be calibrated, while traditional models use default values, sometimes very different from the site specific relationship. Willgoose and Gyasi-Agyei (1995) show that these parameters determine the depth and rate of development of the localised valley erosion. Conceptual Landform Design ERA developed a final landform design in the late 1980s to guide waste dump development and progressive rehabilitation. Some modifications have been made to this design since then. One of the key assumptions with this design is that tailings were rehabilitated both in Pit #1 and the tailings dam (rehabilitation of tailings in the tailings dam has been referred to as the 'above-grade' option in some of the references). Figures 2 and 3 show the contours and how the landform might look following rehabilitation, assuming rehabilitation of tailings in the tailings dam and Pit # 1. (The cover material over tailings and the remainder of the landform is constructed from waste rock.) A later section of this paper describes three other landform designs which have subsequently been developed (see 'Landform Design Alternatives') on the basis of alternative tailings repositories and surface drainage designs. CURRENT EROSION STUDIES One limitation of the input parameters previously derived and applied in the model was that rainfall simulation and wet season monitoring had been carried out largely on bare surfaces of the waste dump and, in some situations, on compacted upper surfaces with no ripping and little or no vegetation. This provided worst case parameters, and as a consequence, worst case erosion scenarios for the landform design due to low infiltration rates. It was recognised and desired by Ranger Mine, that it was important to develop site specific input parameters from wet season monitoring of plots on ripped and revegetated sites on the waste dump, as they would provide a useful comparison with bare sites with no surface treatment. Prior to the wet season fieldwork, which was planned to address this aspect of the model (Evans and Willgoose: 1994), a sensitivity study was carried out to assess the likely magnitude of the effect of vegetation on erosion over 1 000 years (Willgoose: 1995). This assessment was carried out on one of the final landform options for Ranger. The landform evolution model SIBERIA is currently being applied to the Ranger site following the 1995 fieldwork which included gathering of local data from wet season monitoring of

162

Figure 1:

lao

0 0 0

'"

3rd International

21st Anriual Minerals Council of Australia Environmental Workshop

Ranger Mine current layout

'"

ao

'" 0 0

""

0

I~

I

:;

:;

. 0 0 0

'" 0 0

'" 12500N

12000N

SPRAY IRRIGATION 11500N

TAUJNGS DAII IOOOON

9500N

IIBL FLOOD IRRIGATION

Table 1 Overview of erosion and landform design research TIMING OF

CONDUCTED BY ERA

RESEARCH

CONDUCTED BY ERISS

1986/87

1:5 slope construction and wet season monitoring on the northern waste rock dump. Minesoil studies Fire trial on revegetated areas

1987/88

1:3 straight and 1 :3/1 :5/1:8 concave slope - construction and covering with two different rock types, followed by wet season monitoring. Erosion pins also installed across slopes. Auto samplers and depth probes installed adjacent to flumes. Gerlach troughs for bed load monitoring.

1988/89

studies of natural landforms in the region

Monitoring of straight and concave slopes - modifications to aprons to seal with concrete and improve quality of data

LIMITATIONS

• • • • • •

•

•

•

manual sampling of runoff water during storms sampling not carried out over whole slope no delineated plot boundaries no replicated treatments limited range of rainfall events permeable aprons near flumes samplers needed to be activated manually limited range of rainfall events results difficult to apply in terms of selecting specific waste rock materials for covering the whole waste dump

VALUE

•

• •

• • •• • •

• • 1989/90

Monitoring of straight and concave slopes

1990/91

Revegetation research Five year rehabilitation plan introduces landform design assuming tailings rehabilitated in tailings dam

Dry season rainfall simulation of 'cap' site, and 1:5 slope. Wet season monitoring. measurement of key parameters in woodland

1991192

Topsoil trials Revegetation trials Minor modifications to landform designs

Dry season rainfall simulations on ripped plot

measurement of seasonal and storm related trends in solute yield and suspended solids

regular sampling of runoff using auto-samplers more accurate storm related solute and sediment yield relationships erosion pin data clear plot boundaries regular sampling of runoff using auto-samplers more accurate storm related solute and sediment yield relationships erosion pin data additional wet season data (more storms = better data set) different rock types showed different runoff water quality and weathering characteristics flow actuated sampling improved completeness of data set

'as above'

'as above'

•

plot size

•

•

small plots

•

• •

provided data for initial modelling

large range of rainfall intensities comparison of six rip patterns assess temporal changes in hydrology of rip patterns

-

-----_-

TI.MINGOF

CONDUCTED

BY ERA

CONDUCTED

1992/93/94

Revegetation research

Tin Camp Creek monitoring and rainfall simulation

1992/93/94

Revegetation research

Ranger cap and batter site large scale plot monitoring

1993

Ranger cap and batter large scale rainfall simulation

1994/95

Wet season monitoring oftwo vegetated plots on the northern waste dump

1995/96

Development of four alternative landform Burning of wet season . designs and new input parameters on basis of monitored plots 1994/95 wet season monitoring for followed by rainfall application of SIBERIA. simulation

Beyond

VALUE

LIMITATIONS

BY

ERISS

RESEARCH

Refinement of landform designs using model. Streamlining of DIM generation. Application of model at other sites within North Limited, following development of site specific input parameters.

Verification of model using other sites

• • •

• •

•

•

• •

limited range of monitored rainfall events

unvegetated

restricted range of storms difficult to exactly simulate natural rainfall an observer could not be present for all events limited range of intensities difficult to exactly simulate natural rain

• • • • • • •

• • •

influence of long term weathering • on erosion rates not known biological aspects not yet taken into account in model

•

allow comparison between natural and rehabilitated sites provide an understanding of temporal changes in model parameters provided data for comparison of effects of vegetation and ripping useful for assessment of scale effect in modelling can control application of rain controlled data collection for comparison purposes a range of storms can assess effect of vegetation on erosion of WRD under natural rainfall allows direct comparison of effects of vegetation under repeatable data collection conditions allows effects of fire on erosion rates to be assessed ability to model before building landform, must provide improved efficiency when compared with alternative of having to construct large landforms then monitor incorporation of landform design modelling into operations means that landform design is taken into account when mine plans are changed in any way.


165

vegetated sites. This fieldwork provided the data for the development of new parameters (Appendix 3) to be applied to the model (Saynor et al: 1995). EFFECT OF VEGETATION ON LONG TERM STABILITY Using the conceptual design (Figures 2 and 3) as a basis, Willgoose (1995) carried out a preliminary assessment of the effect of vegetation on long term stability of the landform using tabulated values of erosion-vegetation correlations widely accepted in the agricultural community. These erosion predictions were then compared with earlier predictions based on unvegetated sites. The purpose of this study was to provide a quick assessment of the likely influence of vegetation on the magnitude of erosion. Table 2 summarises the findings of this assessment. Table 2:

Comparison of erosion under a range of vegetation scenarios

Max depth erosion Max depth deposition Length longest valley Max depth erosion tops of batters

Unvegetated case

Undergrowth sensitivity case (50% cover)

8.8 m 5.1 m 1600 m 4m

6.5m 2.9m 1050 m 1-2 m

-

Nommal vegetation case (fully developed u'growth & canopy) 5.4 m 2.4 or 2.2 m 900m 0.5-1 m I'll!

'I

The following three sections explain in more detail the predictions summarised in Table 2. Unvegetated Case Figure 4 is the landform before any erosion modelling (with about 15x vertical exaggeration). Figure 5 shows a simulation of the erosion over 1000 years on this landform assuming no vegetation. Extensive valley erosion can be seen and this extends to a depth of 8.8 m. Between these valleys erosion depth is quite small, typically less than 0.1 m. The exact position of these valleys cannot be predicted (Willgoose and Riley: 1993), so the position can only be estimated probabilistically. Factors that can affect the exact position of the valleys include; • spatial variation in settlement and consolidation of the waste rock and tailings, • spatial variation in the runoff and erosion properties, and • minor construction effects which can control the initial runoff pattern, such as wheel tracks and ripping. Undergrowth Sensitivity Case Figure 6 is a simulation of the erosion over 1 000 years on the landform when subjected to the erosion rate of the undergrowth sensitivity case (ie. undergrowth develops to 50% cover during the wet season). As with the unvegetated case, erosion is concentrated into localised valleys. Table 2 summarises how maximum erosion and deposition are intermediate between that of the unvegetated and nominal vegetation cases. Nominal vegetation case Figure 7 is a simulation of the erosion over 1 000 years on the landform with nominal vegetation. As with the unvegetated case, erosion is concentrated into localised valleys but these are no where near as extensive as in other simulations. Table 2 summarises how maximum erosion and deposition vary from the un vegetated case. Overview of sensitivity study The best estimate of the effect of vegetation suggested that the erosion with fully developed vegetation would be 5.8% of the erosion rate of the unvegetated plots. The most significant variable in these calculations was the effectiveness of the 'undergrowth, mostly spear grass, on the

166

Figure 2:


Contours of final landform (assuming tailings rehabilitated in tailings dam)

f .

I

3rd International 21 st Annual Minerals Environmental Workshop

Figure 3:

Council of Australia

167

Current mine layout and artist's impression of final landform

Power Station Retention Pond 2 Ranger No.3 Orebody Ranger No.1 Pit

Retention Pond 1

Tailings Dam

250 500 750 1000 Scale in Metres

Ranger, May 1991

Power Station Retention Pond 2 Ranger No.3

Retention Pond 1

Environment Laboratories

Proposed final landform


168

protection of the surface from the erosive forces of the overland flow. The effectiveness of the canopy to protect the surface from raindrop impact was minimal. If the spear grass fails to develop to a full cover, but only a 50% cover of the surface, then the erodibility of the surface is 20% of that for the unvegetated surface. Definitive estimates of the relative and absolute effect of vegetation were not possible until the results of wet season vegetated plot runoff-erosion experiments were analysed. WET SEASON MONITORING TO DEVELOP INPUT PARAMETERS In November 1994, two large scale erosion plots were constructed on the northern waste rock dump at Ranger (Figure 8). One plot was constructed on a surface ripped, topsoiled and revegetated about 5 years ago (Site 1 - soil site). The vegetation on this plot was typically shrubs (Acacia dominated) and grasses (spear grass dominated). The second plot (Site 2 - fire site) was constructed on a fire trial site where the surface had been ripped and revegetated about 10 years ago and a fire had been introduced when revegetation was young (- 18 months old), to assess the influence of fire on species composition and density. Vegetation on this plot was well established with tall Eucalyptus species, grevilleas and acacias with a slightly less dense understorey of grasses. A total of 78 natural rainfall events were monitored during the 1994/95 wet season between 29 November 1994 and 7 April 1995. Sediment loss and discharge data were collected for the derivation of model parameters. Results indicated that as the wet season progressed there appeared to be a decrease in the total sediment loss. This may be due to an increase in vegetation cover resulting from growth during the wet season. Sediment loss from the soil site (Site 1) was generally greater than that from the fire site (Site 2). This results from the greater vegetation cover on the fire site and the rip pattern. Incorporation of these data and parameters in SIBERIA is in progress and model predictions will be used to further investigate the potential benefit of vegetation cover on long term erosional impacts. Appendix 3 describes how the parameters were derived. LANDFORM DESIGN ALTERNATIVES Four conceptual landform designs were developed from the matrix of options shown in Table 3. These landforms are shown in Figure 9. (Option 1 is the current option). Table 3 Landform design options

Water shedding surface Water retaining surface

Rehabilitation of Rehabilitation of tailings in Pit #1 tailings in Pit #1 and the tailings and Pit #3 dam 1.00 3.00 2.00

4.00

Water retaining surfaces on the final landform have the advantage of minimising runoff, and maintaining high moisture contents in the cover materials over tailings. Thus there are both erosion control and radon minimisation advantages with these options.

11II 3rd International 21st Annual Minerals Council of Australia Environmental Workshop

169

Figure 4 perspective of elevations (starting case,Year 0)

N

~

Figni'e6·

a) Perspective of elevations (undergrowth sensitivity case, Year 1000) b) Plan-of erosion and deposition

Figure 7 a) Perspective of elevations (nominal vegetation case, Year 1000) b) Plan of erosion and deposition

N N

l'

.~

3rd International

170

Figure 8:

21 st Annual Minerals Council of Australia Environmental Workshop

Location of monitoring Sites 1 (soil site) and 2 (fire site) on the northern waste rock dump at Ranger o 100 200m

u ~~"

Ronqer

t)co"""'_

Current research programs are focussed on addressing all of the issues relating to determining Best Practicable Technology (BPT) for tailings rehabilitation. BPT is defined as follows; EPT is that technology from time to time relevant to the Ranger Project which produces the minimum environmental pollution and degradation that can reasonably be achieved having regard to; (a) (b) ( c)

(d) (e) (f)

the level of effluent control achieved, and the extent to which environmental pollution and degradation are prevented, in mining and milling operations in the uranium industry anywhere in the world, the total cost of the application or adoption of that technology relative to the environmental protection to be achieved by its application or adoption, evidence of detriment, or of lack of detriment, to the environment after the commencement of the Ranger Project, the physical location of the Ranger Project, the age of equipment and facilities in use on the Ranger Project and their relative effectiveness in reducing environmental pollution and degradation, and social factors including possible adverse social effects of introducing new technology. Uranium Mining (Environmental Control) Act (1979)

MODELLING DIFFERENT LANDFORM DESIGNS These four design options will be initial conditions to SIBERIA simulations which, using the parameters derived for vegetated and unvegetated conditions, will allow the investigation of the long term erosion properties of these designs, and their relative merits. Insights will be used to further improve the more promising options. An important characteristic to be investigated is the risk of failure of each of the designs. Any design, no matter how good, will have finite pr.obability of failure. The important issue is to balance the risk and cost, a common trade-off in risk assessment exercises. Failure may be defined in many ways but the main one to be used here is the risk of release of tailings (ie structural failure) in 1 000 years. As previously noted, the exact position of the valleys shown in Figures 5, 6 and 7 cannot be predicted deterministically; their position will depend on the random spatial distribution of spoil settlement, runoff and erosion properties. Probabilistic assessments for failure may then include:

171


• •

determining the probability that the depth of erOSIOn at some point on the landform exceeds some critical threshold depth, or the probability that within some region a point will have a depth of erosion exceeding the threshold depth.

Figure 9: Landform designs

Rehabilitation of Tailings in Pit #1 and the Tails Dam

Rehabilitation of Tailings in Pit #1 and Pit #3 M~~.

(~ ~PIT#3

(

c .Ql en (])

o Cl C

C

-

co (]) a::

Conceptual landform design options KEY Drainage divide Flow lines Batters

--

Water Outline of pit#1 and tails dam

172


In this latter case this region may exceed the planar extent of an encapsulated tailings facility and failure indicates that at some point erosion has penetrated the capping layer over the tailings. This probabilistic assessment procedure (Willgoose and Riley 1993) involves Monte-Carlo simulation of a number of scenarios, each of which has random variation of environmental properties typical of that for the waste rock dump. Each scenario is run for 1 000 years and the erosion depths at each point (or any other characteristic that may be deemed to be important) for the scenarios are combined to yield a probability distribution for erosion depth at that point. This probability distribution is evaluated for every point in the rehabilitation and contours of statistical properties can be drawn; ego the probability that erosion depth exceeds some threshold. Willgoose and Riley (1993) and Willgoose and Gyasi-Agyei (1995) demonstrated the insights that can be gained from this procedure using one of the rehabilitation options for Ranger. A general conclusion is that for any given scenario the regions of deep erosion are very localised and separated by regions of low erosion. However, when viewing all scenarios it is apparent that large regions of the rehabilitated landform may exhibit deep erosion, though the probability of deep erosion at any given point is low. The aim of the design procedure is to eliminate designs that show unacceptably high probabilities of catastrophic failure. It is also important that this procedure develops rehabilitation options that when they fail (as they must for an extreme condition), fail in a controlled and non-catastrophic fashion and which for more normal conditions have acceptable behaviour. Thus an important part of the proposed work assessing the four design options with SIBERIA will be to explore post-failure behaviour so that the seriousness of failure can be rationally assessed. This is particularly important for the 'water retaining surfaces' where they have been designed to contain all rainfall received on the surface of the landform. However these 'retention basins' on the surface of the landform may fail causing surface water to overflow during extreme runoff events. A key issue to address is whether the designs will fail such that they are irreparably damaged and prevent containment for future events, or will they fail only to have an impact during that extreme event and not for future less extreme events. Since structures typically fail at their weakest point Monte-Carlo simulation of the random variation of properties will be likely to yield more stringent, but more realistic, design conditions than would occur if random variations were ignored. Incorporation into mine planning Digitising of landform designs as modifications are required can be a tedious process. It is intended that any modifications that are required to designs in the future can be made by environmental or mining personnel directly onto a work station using DAT AMINE, the mine planning software currently used at Ranger. This will enable greater flexibility in planning and a quicker method of regenerating DTMs for input to SIBERIA as needed. Phillips (1991) describes the use of computer aided design to integrate reclamation and mine planning and highlights the direction of this work in terms of providing a direct interface with erosion modelling. O'Reagan et . al (1991) describe the application of computer design of rehabilitated landforms to coal mines in QLD, and empasise the need for an interactive approach to mine planning and landform design. Ideally if erosion modelling can be readily interfaced with mine planning software, design modifications can be rapidly tested and assessed. Related research projects In parallel to the erosion studies carried out in the 1980's were others on waste rock weathering and engineering stability of stockpiles (Milnes et al, 1988). Over the next few months a review will be carried out of the studies completed on long term weathering of waste rock and soil formation processes. This project will provide input data to the model to take account of soil erodibility changes over the very long time frames that the landform evolution model covers. Soil formation and moisture content changes influence the radon emanation rate from waste rock surfaces and current studies are investigating how this material will behave as a cover over tailings in suppressing radon and gamma radiation. Revegetation research has focused on establishment of self-sustaining ecosystems on waste rock surfaces with mature woodland species being introduced with mycorrhizal fungi. While ongoing monitoring of these areas continues, the emphasis over the next few years is on the introduction of native groundcover species. This will be important for long term stability of surfaces. Elevated wetlands will be constructed on the waste rock dump as part of the assessment of alternative landform drainage designs. ERISS are carrying out model validation projects in surrounding areas to improve the reliability of the model.

3rd International 21 st Annual Minerals Council of Australia Environmental Workshop

173

CONCLUSIONS Where design of landforms for long term stability over time frames of 1 000 years or more is required, modelling is a necessary component of the research. At Ranger it is enabling a range of landform designs to be simulated under a range of conditions in terms of vegetation type and cover and soil materials and in the future hopefully, the model will be able to accommodate long term changes in the mines oils due to weathering of waste rock. These improvements in the input data will provide a more accurate model upon which to base long term rehabilitation planning decisions. At the same time the capabilities for rapidly producing Digital Terrain Models using mine planning software will be developed, to provide the flexibility required to make adjustments to the design and have the design tested via application of SIBERIA. REFERENCES Australian Government Publishing Service, Canberra (1982) Code of Practice on the management of radioactive wastes from the mining and milling of radioactive ores East, T J, Uren, C J, Noller, B N, Cull, RIF, Curley PM and Unger C J (1994) Erosional stability of rehabilitated uranium mine structures incorporating natural landform characteristics, northern tropical Australia, in Z Geomorph NF,38,3, 283-298 ERA Ranger Mine (1995) Long term Research Plan and Operating Manual, internal planning report prepared by ERA Environmental Services Pty Ltd for Ranger Mine, July 1995 ERA Ranger Mine (1996) Five Year Rehabilitation Plan (draft) Internal planning report prepared by ERA Environmental Services Pty Ltd for Ranger Mine, January 1996. Evans KG & Willgoose G R (1994) An experimental study on the effect of vegetation on erosion of the Ranger Uranium Mine waste rock dump: A proposal: Internal report 164, Supervising Scientist for the Alligator Rivers Region, Canberra. Milnes A R, Emerson, W W, Richards, G G, Fitzpatrick R Wand Armstrong A B (1988) The long term stability of waste rock dumps in the Ranger Project area, NT Australia - Proceedings, Volume II, environmental Workshop: 79-94. Darwin, September 1988. Australian Mining Industry Council O'Reagan G J, Lowell, A and Roe P (1991) Computer based design of rehabilitated strip mine landforms, Queensland Coal Symposium, Brisbane 29-30 August 1991 Phillips, M (1991) Reclamation - The use of Computer Aided Design to integrate reclamation and mine planning, Queensland Coal Symposium, Brisbane 29-30 August 1991 Saynor, M J, Evans, K G, Smith BLand Willgoose, G R (1995) Experimental study on the effect of vegetation on erosion of the Ranger Uranium Mine waste rock dump. Natural rainfall monitoring data 1994195 wet season, erosion and hydrology model calibration, Supervising Scientist for the Alligator Rivers Region, Internal Report 195, August 1995 Willgoose G (1995) A preliminary assessment of the effect of vegetation on the long term erosional stability of the proposed above-grade rehabilitation strategy at Ranger Uranium Mine, February 95, Open File Report 119, Supervising Scientist for the Alligator Rivers Region Willgoose G and Gyasi-Agyei Y (1995) New technology in hydrology and erosion modeling for mine rehabilitation. APCOM XXV Application of computers and operations research in the mineral industries, Brisbane 10-12 July 1995 Willgoose G & Riley S J (1993) Application of a catchment evolution model to the prediction of long term erosion on the spoil heap at Ranger Uranium Mine. Open File Report 107, Supervising Scientist for the Alligator Rivers Region Willgoose, G, Bras, R and Rodriguez-Iturbe I (1989) A physically based channel network and catchment evolution model, TR 322, R M Parsons Laboratory for Water Resources, Massachusetts Institute of Technology, Cambridge, USA, 464p.

'I

3rd International

174

21st Annual Minerals Council of Australia Environmental Workshop

APPENDIX 1: ERA'S RESEARCH PROGRAM BROAD RESEARCH GOALS Through innovative research and development, apply site-specific best practical technologies to the Ranger mining operations, focussing on: • efficient operation of the mining operation • rehabilitation of mining operations and disturbed ecosystems • protection of the local and regional environment • protection of the workers and the general public r • identification of future land use operations that are compatible with the ecology of Kakadu and the desires of Aboriginal people 1

Mine production

Mine production and management practices that ensure cost-effective operations, minimal environmental impacts, and maximum profits for mining and uranium production. • • • 2

design of operational strategies which provide minimum water and tailings storage volume, maximum water recycle, minimum discharge of contaminants, and efficient construction of the final landform. in situ leaching of low-grade ore stockpiles for maximising recovery of metals and minimising long-term leaching into the environment. strategies for efficient management of the mine water supplies that prevent adverse impacts on the surrounding environment. Tailings

Tailings management practices that ensure efficient deposition, encapsulation and rehabilitation of tailings impoundments.

•

design of appropriate capping systems which ensure cost-effective construction, optimal consolidation of tailings deposits, and stability of the final landform. technologies for effective consolidation of tailings deposits

3

Stockpiles

•

Stockpiles management practices that ensure efficient placement of materials to minimise doublehandling and appropriate stabilisation and revegetation methods. ;; 'I

I

•

• • • 4

technologies for construction of stockpiles, including waste rock dumps, to form final landforms that behave and evolve in a predictable manner in the long term, with minimal environmental impacts, and have a land use compatible with the surrounding National Park. technologies for measuring and monitoring geotechnical, hydrologic and weathering processes within stockpiles, in advance of developing design criteria for the final rehabilitated landform. design principles for the surface shape of the final landform to minimise erosion and maximise capacity for sustainable vegetation. design principles and technologies for broad-scale revegetation of the final landform to produce a self-sustaining, diverse native forest Catchments

Strategies for the effective management of natural processes (geomorphic, hydrological and biological) in the Magela Creek catchment that impact on, or might be affected by, the mining operations. • •

fire management strategies for the effective care of native woodland and mme infrastructure. understand mechanisms of transfer of potentially harmful contaminants from the mine operation into the natural environment, and the development of effective monitoring technologies.


•

175

•

knowledge of wetland processes in natural and constructed water bodies in terms of contaminant movement and impact on water quality and ecosystem diversity. design principles for the construction of small and large scale wetland filters that are compatible with natural wetland processes.

5

Regional environmental issues

Strategies and technologies relating to regional environmental issues (Kakadu National Park) which may have an impact on the mining operations, or the effectiveness and sustainability of agreed end land uses for the mined area. Outcomes will include compatible management practices for feral animals, weeds and fire, maintenance of ecosystems and species diversity; maintenance of traditional foods for Aboriginals; development of tourist and educational potential; and assessment of potential commercial ventures that are compatible with Aboriginal interests. • • • .•

w

1

I·

effective habitat construction and breeding technologies for selected local native flora and fauna species. management strategies for feral animals and weeds. regional fire management systems for the maintenance of natural ecosystems. monitoring of aquatic ecosystems within and outside the mining area.

(From Long Term Research Plan and Operating Manual, ERA: 1995)

'\ ':1 I