Search Documents

Blasting is at first sight a simple operation. Blastholes are drilled in a suitable pattern, primed, charged with explosives, stemmed and finally the blastholes are connected together to give the appro- priate initiation sequence. In practice the action of the explosive takes place in the interior of an opaque rock mass and only the external faces of the rock mass to be blasted can be easily observed. The mechanisms of blasting cannot be observed directly and must be inferred from observations of the free face and meas- urements of particle movement made by sensors grouted into holes in the rock. Explosives react very rapidly, in some microse- conds and any instrumentation of a blast must have an appro- priate response time. In many mines several rock types can be encountered during production. Several different methods of blasting may be necessary to cope with changes in rock types espe- cially if weak rocks are found in the final limits. This also makes it difficult to be sure of getting a large enough volume of rock to perform a series of reproducible experiments. While there appears to be agreement on the mechanism of blasting it is unlikely that a detailed understanding of how all the variables that can be encountered in practice affect blasting. The ease of assessment of the results of blasting depends on the criteria used. It is easy to isolate, measure and count over- size boulders. It is however difficult and costly to get a complete fragment size distribution. Methods used to assess blasting can vary significantly from site to site and the ways in which data are collected, presented and analysed will vary making compari- son of blasts at several locations very difficult. The definition of an optimum blast can also vary with the type of blasting. The muck-pile profile and fragmentation will be different for a dragline to that for a truck and shovel operation. In order to be able to assess blasting the mechanism has to be understood. Th rock and explosive properties considered rele- vant to blasting can then be determined. This paper discusses the generally accepted mechanism of blasting. The on-site observa- tions to assess blasting are-then listed. The mechanism of blast- ing can also be modelled and the use of modelling to optimize hlaetina ie thPn Aiernccarl

Jan 1, 1988

It is a great pleasure to accept the invitation to open this Joint Conference of The Australasian Institute of Mining and Metallurgy and the American Institute of Mining, Metallurgical and Petroleum Engineers. It is the second invita- tion to participate in the Conference that I have accepted. When it was being planned, the President wrote to me to invite me to be one of the speakers in one of the technical sessions. I wrote accepting the invitation, and when my present appointment was announced, there was a nice question about the earlier commitment. The upshot of it all is revealed by the present arrange-., ment: you get me as opener and the more sub- k stantial speech is to be given by Professor Rupert Myers, my friend and former colleague who will not only be able to contribute from the standpoint of an experienced University admini- strator, but will also speak from a professional training and experience much more closely relat- ed to the concerns of this Conference than I possess. Though it is a joint meeting and the speakers will include distinguished figures from North America, I hope that since the venue is Australia, I may be permitted to speak with an Australian emphasis. I have recently had a brief glimpse of part of the Australian mining industry. In the course of a journey through the Northern Terri- tory we-my wife and I-visited Gove, Groote Eylandt, and Tennant Creek. At Gove we saw the remarkable Nabalco development in bauxite and alumina and the very impressive town plan- ning development at Nhulunby, the major man- ganese enterprise at Groote Eylandt, and the gold mining activity at Tennant Creek. For lay people like us, a gold pour was arranged at Tennant Creek and that was a spectacular sight. On earlier occasions, and before my present appoint- ment, I had visited Mount Isa and the Utah fields, and Queensland Alumina at Gladstone. I have not seen the cradle of the modern mining industry, using cradle perhaps loosely, at Broken Hill, nor have I seen the great Pilbara develop- ment, nor Weipa. This is not for want of generous invitations and it is my hope that within the not too far distant future this will be rectified. Then there is the Australian petroleum industry and the great industries which make their products from the minerals mined which I have yet to See.

Jan 1, 1978

By J R BISHOP

Geophysical techniques are being increasingly used down exploration drill holes as deeper and more elusive targets are sought. This paper lists the various methods available and discusses the applications using a number of examples.

Jan 1, 1992

By Y Potvin

Underground mines in Australia are progressively operating at greater depth, sometimes in conditions where mine induced stresses are very high. In such conditions, the seismic activity often becomes intense and violent and in response, mines must develop and implement a sound rock burst risk management strategy to continue operating safely.This paper discusses the latest techniques successfully used in Australian mines to mitigate the risk of rock bursts. In particular, the paper describes how the analysis of data collected from seismic monitoring systems can provide an understanding of the seismicity and be used for quantifying seismic hazards in the mine. Once an adequate understanding and assessment of seismic hazard has been achieved, the associated risk can be mitigated using different layers of protection.Strategically, modifying the mining method, the extraction sequence and/or the mining layout can alleviate the seismic hazard, but these solutions generally cannot be implemented rapidly. In the short-term, seismic risk can be mitigated tactically using an energy absorbing support system or to remove personnel from being exposed to rock burst damage by implementing exclusion zones.

Mar 21, 2011

At the Waikato North Head Ironsand operation, NZ Steel Mining Ltd is using a Bucket Wheel Excavator û shiftable conveyor mining system to excavate Titanomagnetite bearing sand and clay. The system utilises two Bucket Wheel Excavators with a two level shiftable bench conveyor and common trunk conveyor to deliver the material to a processing plant. The system was commissioned in November 1984 and has proved itself to be a high capacity, low cost method of excavating. Techniques of handling high sand faces have been developed and the conveyor layouts are going through a process of change to cope with the variable geo technical conditions and remain within: the characteristics of the excavators. The safety requirements for operating the system have been of a high standard and the experience has increased the confidence of the Company and the Mines Department Inspector's, for the first Bucket Wheel Excavator Mining system in New Zealand.

Jan 1, 1986

By B McFadyen, Y Potvin, K Woodward

The Australian Centre for Geomechanics (ACG) held two workshops in 2019—one in Australia and one in Canada—to discuss processes and methods currently used in the mining industry for stope design and reconciliation. In practice, this work strives to optimise stope performance and aims to maximise ore recovery (minimise underbreak), while minimising dilution (overbreak). One of the key findings from the workshops was that current stope design practices, almost exclusively, rely on some versions of the empirical Stability Graph methods developed in 1980s. Although robust and useful at the feasibility study stage, there are a number of inherent shortcomings with this approach. The Stability Graph’s limitations first lie in that it cannot assess underbreak, which is a very important measure of performance. Furthermore, the method doesn’t consider a number of factors which often have a critical influence on stope stability, for example, drilling and blasting or the presence of faults. Given the current industry standard practice, there is a great opportunity to reinvent the stope design process to optimise performance. A finding of the workshops was that every mine quantifies stope performance in terms of volume reconciliation with a focus on dilution and ore losses, and reconciling metal tonnes recovered from the mill. Many mines further investigate stope performance by looking at individual stope faces and the causes for the deviations between the planned stope shape and the measured cavity. Recent research at the ACG has shown that investigating overbreak and underbreak at a higher resolution is a powerful way to identify the root causes of overbreak and underbreak. Finer scale assessment is achieved using an octree analyses, which effectively divides the relevant stope volume into cubes on the scale of a metre and produces many tens of thousands of data points. The octree data structure and subsequent analysis underpins a data-driven stope design methodology. The ACG is currently developing a design approach based on multivariate modelling methods that can account for the many variables that influence stope performance and forecast stope overbreak and underbreak results. Back-analysing stopes based on octree reconciliation results provides a rich database for training the model. Once the model can achieve reliable forecasting of overbreak and underbreak, an interactive drill-and-blast design will allow for the assessment of different firing strategies to optimise stope performance.

Mar 15, 2021

A novel approach to validate size distributions obtained from bubble sizing techniques is introduced. The experimental rationale consists in verifying whether the output features the anticipated bi-modality when two uni-modal populations (ca. 0.5 and 5 mm) are mixed in a bubble column. The degree of deviation is quantified by comparing the calculated and real proportions on a volumetric basis. A sampling-followed-by-imaging technique was used to proof the concept. The results showed the anticipated bi-modality in which the real and calculated proportions corresponded with a deviation of less than 15 per cent, considered good given the severity of the test. However, the deviation was correlated to the superficial gas velocity (Jg) in the column and was consistently against smaller bubbles. Possible sources of bias are suggested.

Jan 1, 2005

By K Biegaj

During the author’s 25+ year career in mine management in various parts of the world (mostly in Australia), the author has been privileged to live and work in mining towns (not FIFO) for more than 20 years. And has personally observed other people from the mining industry carrying out work at home in an unsafe manner, eg pushing lawn mowers whilst wearing thongs or performing electrical jobs whilst standing on aluminium ladders, as well as many others. It is common that injuries sustained at home whilst on break are not reported upon arrival back on a mine site. A large number of compensation claims for injuries at work, which in fact had been sustained at home, come at substantial extra costs to mining companies.

May 5, 2022

By P Joyce, D R. Cooke

The Dinkidi Cu-Au porphyry deposit is situated in the Caraballo Range of North Luzon, Philippines. Although comparatively small, Dinkidi has unusually high gold grades with drilling by Climax-Arimco Mining Corporation delineating a 110 Mt orebody at 1.2 g/t gold and 0.5 per cent copper. The magmatic-hydrothermal evolution of Dinkidi records a distinctive switch in mineralisation from an early orthoclase-dominated Au-rich assemblage, typical of alkaline porphyry systems, to a later quartz stockwork Au-Cu rich assemblage typical of most calc-alkaline to shoshonitic porphyry systems. The early orthoclase-dominated stage is associated with the intrusion and crystallisation of a variably-textured clinopyroxene-syenite dyke that contains high-grade (5 - 20 g/t) gold mineralisation. We consider this dyke to be a Cu-Au porphyry analogue to the pegmatites found in granitic systems. Dinkidi is hosted within the Early Miocene Didipio Igneous Complex. This alkaline igneous complex is comprised of early unmineralised gabbros (pyroxene-plagioclase cumulates), clinopyroxenites (pyroxene cummulates), diorites and monzodiorites that have been intruded by a large, weakly mineralised monzonite pluton. The Dinkidi deposit is located at the southern margin of the monzonite pluton. Mineralisation is hosted within a composite monzonite stock. Within the stock, an 800 metres by ~200 metres (plan section) elongate monzonite stock (Tunja Monzonite) predates and hosts the bulk of the Au-Cu mineralisation. This has been intruded by the thin (10 to 30 m wide), highly mineralised variably-textured magnetite- and clinopyroxene- phyric syenite pegmatite (the Balut Dyke). The youngest phases of the composite stock are leucocratic quartz-monzonite porphyry (Quan Porphyry) that grades texturally, in its core, into a post-mineralisation crystal-crowded quartz-syenite (Bufu Felsite). The late-stage Bufu Felsite is inferred to be the source of much of the Cu-Au mineralisation. The Didipio Igneous Complex has intruded a sequence of Late Oligocene alkaline andesitic lavas and volcaniclastics and overlying potassic dacitic lavas. Whole rock geochemical analyses indicate that the intrusions and the potassic dacite lavas in the surrounding region are co-magmatic. Five main stages of hydrothermal activity are recorded at Dinkidi. An unmineralised K-silicate (biotite-magnetite) alteration assemblage formed during the intrusion of the Tunja Monzonite stock. The diorites surrounding the stock display intense selectively pervasive biotite and magnetite alteration and are cut by an extensive network of orthoclase veins and monzonite dykes. The orthoclase-dominated assemblages of Stage Two carry substantial high-grade Au mineralisation and are associated with the intrusion of the Balut Dyke. The emplacement of this unusual dyke is associated with emplacement of pegmatitic clinopyroxene-orthoclase-magnetite¦sulphide dykes, localised orthoclase flooding along the margins of the dyke, and subsequent emplacement of an extensive orthoclase-actinolite-bornite¦chalcopyrite-fluorite stock-work, and development of widespread selectively pervasive actinolite alteration. The main stage of Au-Cu mineralisation at Dinkidi is intimately associated with the intrusion and initial crystallisation of the leuco-monzonite porphyry (Quan Porphyry) and emplacement of a quartz stockwork followed by carbonate-sulphide deposition and intense vein-related sericite and carbonate alteration assemblages. The quartz stockwork formed synchronous with the deposition of the æBugoy QuartzoliteÆ, a very coarse grained mosaic of quartz-orthoclase-perthite along the walls and top of the crystallising intrusion. Stage 3 was terminated by crystallisation of the distinctively vuggy Bufu quartz-syenite phase in the core of the intrusive, with associated antiperthite veins, and the subsequent explosive brecciation of the roof of the intrusion, forming a quartz-fragment breccia from the Bugoy Quartzolite. Retrograde metasomatism during Stage Four was intimately associated with the emplacement of extensive, late-stage carbonate, silica and zeolite veins and widespread selectively pervasive carbonate alteration. The last hydrothermal stage at Dinkidi, Stage 5, is comprised of strong selectively pervasive clay alteration and is associated with fault-related brecciation of the Dinkidi stock.

Jan 1, 1999

By F Sandrone, J-P Dudt, J Zhao, E Paltrinieri

Tunnel boring machines (TBMs) excavation is nowadays more and more considered a valuable alternative to conventional tunnelling. As a matter of fact, thanks to its advantages such as greater rate of advance and safer working conditions, the application of TBMs has increased significantly in recent years. Despite the possible excellent performance of TBMs in favourable ground conditions, the advance rates could be significantly slowed down or even obstructed in limiting geological situations. This obviously traduces in potentially longer tunnel construction times and higher costs and, therefore, in a preference for conventional method. One of the major geological hazards encountered during tunnel construction, is related to the presence of fault zones or heavily jointed and fractured rocks. The most common geotechnical problems resulting from tunnelling in these difficult conditions include the instability of the excavation face, instabilities of side-walls producing large overbreaks on the tunnel profile, excessive deformations in case of squeezing grounds, frequently changing stresses in terms of magnitude and direction, high water inflows, etc. The decisions aids for tunnelling (DAT) is a model and a computer code that can be used to evaluate the influence of the uncertainties related to both the geological/geotechnical conditions and the construction process on final tunnel construction time and costs. In this framework, the DAT is used for assessing the effects that challenging ground conditions may have on tunnelling operations in terms of costs and delays. This is done by simulating the tunnel construction considering several geological profiles. The aim of the study is comparing different excavation methods (ie drill and blast versus TBM) in changing ground conditions along the tunnel alignment ranging from good rock, to very poor and fractured formations. The results show the impact, in terms of changing construction time and cost, of the excavation method versus the ground conditions.CITATION:Paltrinieri, E, Dudt, J-P, Sandrone, F and Zhao, J, 2014. Comparison of excavation methods in different ground conditions with the decision aids for tunnelling, in Proceedings 15th Australasian Tunnelling Conference 2014 , pp 573–580 (The Australasian Institute of Mining and Metallurgy: Melbourne).

Sep 17, 2014

By W Ilahi

Underground metalliferous mining methods often result in the formation of vertical openings to facilitate the extraction of orebodies. These development and production techniques can result in the creation of vertical open voids, which include stopes, orepasses, ventilation raises and other excavations. There are risks associated with the use of mobile plant in the vicinity of these voids, especially during backfilling and when bunds are constructed at the edge of a stope or void. These voids can present unacceptable risks if not adequately managed.

May 5, 2022

S. A. Tomich (Consulting Geologist, Perth) : I would like to commend Mr. Haycraft on his presentation on the economic geology of the North End of Kalgoorlie. This is an intriguing area which is in the Kalgoorlie field but not of it, inasmuch as it is separated spatially from the main Golden Mile by the Golden Pike Fault. Mr. Haycraft has stressed the differences-the mineralogical and structural differences of economic significance between the two areas.I would like to ask whether he is satisfied that the lithological sequence in the North End is the duplicate of that in the main Golden Mile area irrespective of the postulated effects of erosion, and more specifically whether there is any chance that the quartz-dolerite-greenstone of the North End may be in any way dissimilar from the quartz-dolerite-greenstoneof the Golden Mile which is tellurude bearing.J. A. Haycraft: In relating the quartz-dolerite in the Hannan's North-Mount Charlotte area back to the Golden Mile, we find some quite close similarities. Although the rocks are similar they are not necessarily one and the same.At Mount Charlotte we have three varieties of dolerite- a coarse-grained gabbroic dolerite on the eastern side of the orebody, the host rock medium grained and on the western side, a fine grained dolerite. Further west we have an amphibolitic variety of quartz dolerite.Recent drilling south of the Golden Pike fault on the Enterprise No. 20 level had indicated a remarkably similar sequence of rocks. We drilled west through a wide section of medium grained rocks into a fine grained rock and then into amphibolitic quartz dolerite before intersecting the Golden Pike fault. Williamstown dolerite was located west...

Jan 1, 1965

By J Liu

In situ coal gasification is a method of extracting a high proportion of the energy value from coal without the need for mining. In situ conversion of coal to combustible gases is achieved by drilling wells into the seam, injecting gasification agents and initiating combustion. The product gases are then extracted to surface using production wells. These gases are clean energy sources for power generation, heating or synthetic fuel production. In situ coal gasification is of particular interest with respect to its lower environmental impact, potential application to otherwise inaccessible energy resources (ie offshore, deep deposits or beneath protected lands) and low capital intensity. Practical application of this clean technology is restricted primarily by lack of satisfactory quantitative descriptions of the processes involved, in particular the hydromechanical processes.In this study, three characteristic physical zones of an in situ gasifier are identified as the rubble zone, combustion front and virgin coal zone. Coupled mechanical processes of coal consumption and overburden collapse are used to describe the permeability profile in the rubble zone. Additionally, the redistribution of overburden load associated with coal consumption is accounted for with the formulation of a stress (compressive) dependent permeability profile, which is applied to the virgin coal zone. The model was implemented into and solved by FEMLAB, an advanced computer simulation tool. The flow process, according to DarcyÆs Law, is subsequently described in terms of permeability dependent injectivity. Injectivity is an indicator of flow velocity, which has been identified as a crucial factor in determining the commercial value of product gases from in situ coal gasification. The stress alteration to coal permeability was found to have a dominating influence on the flow process. This finding implies that a greater understanding of the in situ coal burning induced deformation of surrounding rocks would be of significant value in assessing the economic viability of applying this clean technology to potential sites.

Jan 1, 2004

By Harper F

This paper outlines the authors' views of the methodologies which can be employed to value mineral properties at different stages of development for the purposes of project financing, mergers and public floats. This paper discusses the key issues in Discounted Cash Flow (DCF) analysis including the importance of data verification and the shortcomings in DCF analysis. Stockmarket valuation measures and valuation methodology for exploration and pre development projects are examined through examples and pitfalls are identified. The use of valuations in corporate transactions and their strategic value is also discussed.

Jan 1, 1994

By Hermulheim W

In the Federal Republic of Germany hard coal is mined underground only. The coal mines reach down to 1 500 m deep, recently having an average depth of 900 m. The commonly used winning system is long- wall mining with face lengths up to 250 in. The gate roads, which separate the seam into single panels, have an average length of one kilometre with a cross- section up to 20 m2. These single entries are usually supported by yielding steel arches. The installation of steel arches can hardly be mechanized. The choice of heavier support, caused by increasing depth and roadway cross-section, results in high cost for labour and material in gate roads. Furthermore, the heavy steel arch is acci- dent-prone, not only during installation but also during transportation and ship- ment. Rock bolts combined with a suitable lagging offer a way of mechanizing support work in headings. The aim is to decrease cost and to increase mine safety and pro- ductivity. Over the last ten years, 50 km of gate roads have been reinforced by a combination of rock bolts and weld mesh. Nevertheless, there are still differing opinions on the feasibility of rock bolting in gate roads. In Germany a long tradition and experience with steel arch support exists, whereas the function of an anchorage is a closed book to miners, preventing this new technique from spreading more quickly. Advantages and shortcomings of both steel arch support and rock bolting are compared. The experience with rock bolting up to now is reported. Possible improve- monts arm discussed.

Jan 1, 1986

By Mutton B. K

Hilton Mine is being developed as an integrated component of the Mount Isa operation, to achieve continuity of silver/lead/zinc production well into the next century. The mine has a chequered history from its discovery in 1947 through several periods of intense development activity, each of which has been curtailed due to the further development of the Isa Mine operation. This era has now passed; to sustain increased production rates at Mount Isa, there is a requirement for Hilton ore in 1989. The decision was made in 1983 to accelerate the programme of Hilton development by installing skip hoisting facilities in the main service shaft (P49), to facilitate the development of the mine to pre-production status. The Hilton development plan provides for the mining of selected orebodies, a broadening of the underground exploration drilling programme, and the establishment of significant mine infrastructure. Several technical challenges have been identified as important. The Hilton orebodies, although occurring in the same rock type as Isa Mine, are more faulted, jointed and intersected by intrusive dykes. The ore is more variable in terms of ore continuity, boundaries, grade and metal distribution and the mineralogy more complex. As a consequence the development of this project consists of a co-ordinated programme of mine exploration and development, rock mechanics investigations and metallurgical research and development work. Production will commence in 1986 at a rate of 1000 t/day using mechanized cut and fill techniques. It is planned to introduce sublevel open stoping and by 1996/97 the production rate will be 2.5 million t/yr.

Jan 1, 1986

By Trois P

A fully instrumented and bench-mounted diamond drilling machine, purpose designed and built, has been used for investigating the effect of aqueous solutions of various chemical compounds when used as drilling fluids. The rocks investigated were andesite, basalt and granite, the chemical compounds were alkylamines, with various hydrocarbon chain lengths, aluminium chloride and dioxane. Rotation speed seems to be one of the major factors affecting the emergence of mechanochemical effects as the results obtained with aluminium chloride clearly indicate. Penetration rate is either unaf- fected or enhanced by the presence of certain compounds, such as aluminium chloride. On the other hand, penetration rate diminishes when the drilling fluid contains dioxane and exhibits an apparently unpredictable behaviour with alkylamine solutions, although it appears that an increase in hydrocarbon chain length usually equates to an increase in penetration 1. Full Professor, Mining and Mineral Dressing Department University of Cagliari - Italy 2. Associate Professor, Mining and Mineral Ikea- sing Department University of Cagliari û Italy rate. Rock texture seems to have a pronounced effect on test repeatability. With hexadecy- lamine acetate solutions in concentrations ranging from 10-5 to 5 .10-4 molò dm 3 and 41.88 radòs-1 rotation speed, 46 % increments in penetration rate of andesite were observed, with respect to water, whereas penetration rate decreased in basalt by as much as 60¦%o when a dioxane solution was used as drilling fluid. An interpretation of the observed effects is proposed in the light of current knowledge of solid state physics and physical Chemistry.

Jan 1, 1992

A short overview of the research program of SveBeFo, Swedish Rock Engineering Research, which was founded in 1993 is first given. Then four recent and ongoing projects in rock blasting and explosives engineering are presented in greater detail: 1. Burn Equations for Commercial Explosives. 2. Blast Damage in Contour Blasting. 3. The Role of Gas Pressure in Pit Wall Damage. 4. Fragmentation of Aggregate Rock. The goal of the bum equation project is to be able to control the energy release from explosive charges to the desired part of the expansion curve of the reaction products under various rock conditions. SveBeFo's work has reached the stage where a burn equation for a lab emulsion has been established which can model the influence of charge diameter on the VOD and detonation front curvature in a cylindrical charge quite well. This work now continues with other emulsion recipes. In the blast damage project the goal is to develop techniques and recommendations on how to minimise the blast damage in construction blasting. The project has gained extensive practical knowledge from the blasting of more than 200 holes in a granite quarry under controlled conditions. The cracks in the rock surrounding the contour holes have been cut open and measured. The influence of factors like charge concentration, decoupling ratio, hole and charge size, VOD and time delay in initiation have been investigated. A simultaneous initiation of a group of contour holes with EDD detonators is quite superior to single hole blasts, even when they are timed as closely as 1 ms.

Jan 1, 1995

In order to satisfy environmental and mine operational requirements it is becoming increasingly important to minimise ground blast vibrations at certain locations. These locations are often in soil or on large resonant structures such as grain silos, highway bridges or even the walls of an open pit. A soil monitoring exercise was carried out to evaluate the common practice of coupling vibration detectors to the ground using spikes. At a series of locations in an operating quarry the spiked array response to single blastholes was compared with that of an accelerometer array attached to a calibrated embedded mount. The results clearly showed that the spiked arrays over-estimated the true ground vibration by 46.5 per cent on average, with a maximum observed error of 99.2 per cent. These errors are considered too large for compliance monitoring where accurate vibration levels are required for comparison with prescribed limits. Although spike coupling is convenient, its use for compliance monitoring should be viewed with caution. Unfortunately, however, many manufacturers supply their detectors as spiked arrays. A structural monitoring exercise was also carried out to evaluate the vibration response of a grain silo, highway bridge and a pit wall to the initiation of single blastholes. In this case accelerometers and standard geophones were firmly attached to the bridge which resonated at low frequencies. The results clearly show that, due to their limited low-frequency response, the geophones typically measured only half the true vibration level as measured by the accelerometers.

Jan 1, 1995

By Schmincke H-U

The Pleistocene basanite-tephrite Rothenberg cone complex was built by an alter- nation of strombolian and phreatomagmatic phases of explosive eruption. Two main centres, on a one km long NNE-trending fissure, were active during its construction. The strombolian deposits from three discrete facies corresponding to the crater-fill and inner wall and outer wall environments on the cone. The outer wall facies consists of an alternation of 1. 'normal' coarse-grained, relatively well- sorted lapilli and bomb beds, and 2. finer-grained poorly-sorted lapilli beds. Evidence suggests that the former was pro- duced by open-vent lava fountaining, with balanced magma supply and discharge rates, and the latter are the products of eruptions during partial blockage of the vent, by collapse of wall-rock, and formation of a degassed crust on the magma. The resulting explosions there- fore ejected a mixture of fresh actively ves- iculating magma, older partially degassed lava and material recycled from the wall of the cones. The phreatomagmatic deposits are an alternation of moderately-sorted fall beds and thinner ash-rich pyroclastic surge and rain-flushed units. The clast populations are much more diverse than strombolian ones, including abundant wall rock fragments from 0-100 m depth. The juvenile (magmatic) particles range from highly vesicular scoria to dense 'bread-crusted' clasts, and contain abundant small wall rock inclusions. The phreatomagmatic phases are products of explosive interaction of small batches of magma with shallow groundwater. Vesiculation of this magma played a small but significant role throughout the fragmentation.

Jan 1, 1987