Search Documents

By Donald F. MacDonald

THE cement plant nearest to Nova Scotia is not far from Montreal. Cement is a heavy product which, because of its relatively low value, cannot stand a high freight charge without unduly enhancing its cost. This is one reason why cement costs are high in Nova Scotia. The regular retail price per bag (8772 lb.) in Antigonish is 85 cents, which would be $3.40 per Canadian barrel of 350 pounds, or $3.65 per American barrel of 376 pounds.It is said that the lowest wholesale price in Halifax is 65 cents per sack, which would be $2.60 per Canadian barrel or about $2.80 per American barrel. According to U.S. Bureau of Mines statistics, the average factory price for cement in 1937 was $1.48 per American barrel, which is practically the same as the value per barrel indicated in Dominion statistics for the Canadian output during that year. The difference between $3.40 and $1.48 is $1.92. It is evident that the average user of cement in the Mari-times pays a stiff premium because Maritime cement raw materials have not been developed. Rock-wool, the manufacture of which is rapidly increasing in Ontario and the United States, is much needed for building insulation in the cold climate of Nova Scotia.

Jan 1, 1939

By Randolph A. Pax, Brian Cornish

The operation of grinding mills in an ore processing circuit is supported by monitoring the input and output variables. These can include the feed ore size distributions, ore feed and water addition rates. The outputs can include the discharge undersize density, oversize tonnage rate and the P80 of the discharge. The grinding mill performance is judged by the power draw, recycle rates and discharge grind size. Conditions inside the mill are inferred from loads cells or bearing pressures and past histories. An overall noise value is sometimes used to evaluate the load as well as the occurrence of excessive impacting of steel balls onto the liners/lifters of the mill. This paper presents results of the internal behaviour of a mill using non-contact stationary microphones located close to the mill outer shell (Pax, 2012). The noise generated by a mill can be attributed to a number of events that occur inside these mills and suitably analysed can provide real-time determinations of in-mill behaviour and can provide powerful insights into the operation of the primary comminution tumbling mills. The acoustically determined signals including a specific 'steel' signal, water hold-up, integrated area together with a classification of impact events. Specific results from case studies will show information of the trajectories inside a mill, steel to rock ratio as well as rheology effects. By incorporating a self-consistent understanding methodology for the in-mill processes from an array of microphones, the information can be used to diagnose the dynamic conditions inside a mill.

Jan 1, 2016

By A. S. Prokopovitsh

A spectrophotometric method for the determination of small amounts of copper in tungsten metal powder using the principle of solvent extraction has been developed by the Federal Bureau of Mines. The metallic sample is dissolved in hydrogen peroxide, tartaric acid added, the pH adjusted, and the excess peroxide destroyed by boiling. Copper is reduced to the cuprous state, and the reagent neocuproine (2,9-dimethy1-1,10-phenanthroline) is added to form a colored complex with the copper. This colored complex is then extracted into chloroform and the absorbance of the separated chloroform layer measured on a spectrophotometer. Copper contents as low as 1 p.p.m. (parts per million) can be determined. The method is limited to finely divided tungsten powder. The interest of the Bureau of Mines in developing more sensitive methods for the determination of trace impurities in high-purity tungsten prompted an investigation of the possibility of using solvent extraction-colorimetric methods. The organic reagents now commercially available make possible many extremely sensitive colorimetric methods for the determination of metallic ions. Extraction of the colored complex into a small volume of organic phase can be used to further increase sensitivity (4) .4/ This general approach has the added advantage that no special equipment beyond that normally found in an analytical laboratory is required.

Jan 1, 1961

By A. Roca, G. Ríos, I. Ruíz, M. Cruells

Flash Smelting Furnace (FSF) hot dust-laden is cooled in a waste heat boiler (WHB) and then dedusted in electrostatic precipitators (ESP’s) before being treated, together with the Peirce Smith Converters off-gases, in the sulfuric acid plants for SO2 removal. FSF dust is sulfated by SO2 and O2 in the WHB. Typically, 70% of the dust from the FSF drops in the WHB while the remainder is caught in the ESPs. The collected dust is recycled to the FSF for Cu recovery. Some smelters have developed hydrometallurgical treatments to recover Cu from the dust; impurities, such as As and Bi, are bled as leach plant residues. This paper presents a study (laboratory scale), of leaching the FSF dust, generated in Atlantic Copper Smelter using the weak acid bleed from the gas scrubbing/cooling section (GSCS) of sulfuric acid plants. Leaching was performed with 0.5 mol L-1 sulfuric acid concentration, 60ºC, and a solid/liquid ratio 1/3. After leaching, H2O2 was used for arsenic and iron oxidation. pH was adjusted for precipitation of arsenic and iron as scorodite or as scorodite and arsenical jarosite. The precipitation was carried out at 95ºC during 3-6 h. Liquids after precipitation are suitable for subsequent treatments of electrolysis. INTRODUCTION The copper smelter dusts are usually recirculated totally or partially to the flash furnace. This practice increases the amount of impurities feeding the furnace. Some authors have developed processes for treatment of these materials for copper and zinc recovery and for arsenic, antimony and bismuth inertization. The processes for the treatment of copper smelter dusts can be pyrometallurgical (Yu, 1987; Carter, Vance, Aldridge, Zaw, & Khoe, 1994; Fu, Jiang &Wang, 2000), hydrometallurgical, or mixed processes (Yin et al, 1992; Mulale, Mwema, & Mashala, 1999; Gorai, Jana, & Khan, 2002). Hydrometallurgical processes have been extensively studied and, in some cases, applied to an industrial scale. The most important chemical reagent proposed for leaching the dusts has been sulfuric acid (Roman-Moguel, Plascencia, Pérez, & Garcia, 1995; Gabb, & Evans, 1997; Larouche, 2001; Morales, Cruells, Roca, & Bergó, 2010; Cáceres, Cruells, Morales, & Roca, 2012; Ecometales, 2013; Yang, Fu, Liu, Chen, & Zhang, 2017). Processes developed in alkaline media, and in some cases with the presence of sulfide, have also been carried out (Gabler, & Jones, 1988; Morales, Hevia, & Cifuentes, 2009; Alguacil, Garcia-Diaz, Lopez, Rodriguez, & Alonso, 2013), as well as by using a bioleaching process (Massinaie, Oliazadeh, & Seyed Bagheri, 2006; Bakhtiari, Atashi, Zivdar, Seyedbagheri, & Fazaelipoor, 2011).

Jan 1, 2019

IN the construction of a set of ordinary mine plans, two axes or coordinates are used; also, two different projections are essential viz., a plan (horizontal projection) and longitudinal section (vertical projection). In isometric projection three axes are employed-one axis vertical and the remaining two inclined to it at angles of 120°, also making an angle of 120° with one another, as shown in Fig. 1. The two inclined axes are used to plot the horizontal projection, while altitudes are plotted on the vertical axis. As an illustration, consider the ease of a simple object, such as a box. To show all the dimensions in the usual manner requires two drawings-a plan and elevation, as in Fig 2.

Jan 1, 1919

By Nader Golchinfar

The V-shaped breakout zones around excavation boundary have been captured by a variety of numerical methods such as FEM and DEM methods and failure criteria such as linear Mohr-Coulomb and non-linear Hoek-Brown failure criteria. Diederichs (2007) used Hoek-Brown parameters in his spalling parameter approach; however, it is convenient to use the Mohr-Coulomb failure criterion in numerical modeling because the material parameters can easily be determined from triaxial tests. It is shown in this study that when peak and residual strength parameters as well as dilation angle are properly defined, the brittle Mohr-Coulomb failure criterion can be used to model depth of failure near underground excavations satisfactorily.

Jun 1, 2012

By V. B. Dale

This report is one of a series covering the mineral resources of the Nation. It briefly describes most of the known tungsten deposits in Gila, Yavapai, and Mohave Counties, Ariz. Many of the descriptions are taken from the reports of previous examinations by Bureau of Mines engineers, and from Defense Minerals Exploration Administration files. Where a previous examination had been made, and an inquiry disclosed that little or no work had been done since that time, pertinent information from the report was used in this paper with acknowledgment of the source. However, the majority of the descriptions is from field examinations by the author. In the past 3 years (July 1956-May 1959) only necessary annual assessment work was conducted on nearly all the properties described in this paper. There was no active mining on any of the properties examined.

Jan 1, 1961

By Wilfred Walker

Much is known of the geography of the Alpine orogeny because it has now ended and all geosynclines are in their terminal, cordilleran form. The principle of sea floor spreading, continental drift, and mantle cells has gained general acceptance, but the number and position of the cells, even during the Alpine orogeny, is controversial. The oceanographers have demonstrated the uprises and in this paper it is suggested that downflows are marked by island arcs in the oceans and geosynclines (now cordillera) in the continents, for a total of about a dozen cells. Major mineral deposits associated-with the Alpine downflows are porphyry copper, molybdenum, nickel, chrome, lead-zinc-silver, gold, mercury, tungsten, and tin. Amore precise environment is known for some of these metals, for example, nickel, chrome, and mercury in flysch (which flanks mantle downturns), Ten major orogenies have taken place from 3500 m.y. to the present, Their re¬construction and metallogenesis is the subject of a continuing study.

Jan 1, 1971

By Tracy C. Jarrett

In making gold alloys possessing certain physical properties, such i~nportant factors as corrosion resistance, color, hardness and melting points must be considered. All of these requirements depend directly upon the composition. It is possible to obtain the same colored gold alloy by using different compositions, but in doing so one may bring in some undesirable property. Other investigations have been carried on in this field by F. E. Carter,' E. M. Wise12 and L. Sterner-Rainer2 To make a complete study of the effect of composition upon these properties, a systematic investigation was carried on involving a range of copper content from 85 to 65 per cent. Elements such as zinc, silver and nickel were used to replace in part the copper content. The gold content in all the alloys remained constant. Only 10-karat gold alloys of those containing 41.6 per cent gold will be considered. All of the alloys were made in an induction furnace. Each alloy was cast in a 3 by 1/4 by 3/4-in. ingot. The method of obtaining the melting points and making color determinations was described in a previots paper.4 The Rockwell B scale with the 1/16-in. ball and 100-kg. weight was used for all hardness determinations. The submersion of test pieces in concentrated nitric acid is one of the standard corrosion tests used in the optical trade. A test piece in the 50 per cent cold-worked condition, % by 2 by 0.045 in., is blanked from a skip using a die, and the weight loss per square inch of surface is taken as a means of comparison. The time for this test is 2 1/2 hr. of actual submersion of the test picce in the acid. This standard test was applied to all of the 10-karat gold alloys discussed in this paper. The tensile tests were made upon strips of the dimensions shown in Fig. 1. The test pieces after blanking were annealed at 1300" F. for 20 min. and cooled in air. Thc "drop of the beam" method was used in determining the yield point, as there was a very marked "drop" when the yield point was reached.

Jan 1, 1940

By Tracy C. Jarrett

In making gold alloys possessing certain physical properties, such i~nportant factors as corrosion resistance, color, hardness and melting points must be considered. All of these requirements depend directly upon the composition. It is possible to obtain the same colored gold alloy by using different compositions, but in doing so one may bring in some undesirable property. Other investigations have been carried on in this field by F. E. Carter,' E. M. Wise12 and L. Sterner-Rainer2 To make a complete study of the effect of composition upon these properties, a systematic investigation was carried on involving a range of copper content from 85 to 65 per cent. Elements such as zinc, silver and nickel were used to replace in part the copper content. The gold content in all the alloys remained constant. Only 10-karat gold alloys of those containing 41.6 per cent gold will be considered. All of the alloys were made in an induction furnace. Each alloy was cast in a 3 by 1/4 by 3/4-in. ingot. The method of obtaining the melting points and making color determinations was described in a previots paper.4 The Rockwell B scale with the 1/16-in. ball and 100-kg. weight was used for all hardness determinations. The submersion of test pieces in concentrated nitric acid is one of the standard corrosion tests used in the optical trade. A test piece in the 50 per cent cold-worked condition, % by 2 by 0.045 in., is blanked from a skip using a die, and the weight loss per square inch of surface is taken as a means of comparison. The time for this test is 2 1/2 hr. of actual submersion of the test picce in the acid. This standard test was applied to all of the 10-karat gold alloys discussed in this paper. The tensile tests were made upon strips of the dimensions shown in Fig. 1. The test pieces after blanking were annealed at 1300" F. for 20 min. and cooled in air. Thc "drop of the beam" method was used in determining the yield point, as there was a very marked "drop" when the yield point was reached.

Jan 1, 1940

By W. J. Taylor

SOME eight years ago I became interested to a considerable extent in one of the well-known deposits of sulphury iron ore in New Jersey, and, as a consequence, soon became interested in the subject of the elimination of the sulphur. On investigation I found that no department of pig-iron manufacture had received less attention than that of roasting or calcining ores for the elimination of sulphur, carbonic acid, water, etc. My first devices to eliminate sulphur were naturally in the lime-kiln line, with solid fuel mixed with the ore, in the same way as common lime is burnt with small coal. While this process will succeed tolerably well for carbonates or calcareous ores, even though they contain sulphur as sulphates, I soon found it would not answer for our magnetites, which are very hard and dense, carrying from 2 to 5 per cent. of sulphur as pyrites, and become very sticky as soon as any attempt is made to roast them at a high heat. Some laboratory experiments developed the fact that the nearer the gaseous current surrounding the hot ore approached in composition pure atmospheric air the more rapidly the sulphur was oxidized, while in an atmosphere composed largely of carbonic oxide, carbonic acid, and nitrogen, it seemed almost impossible to reduce the sulphur more than one-half, which was a confirmation of the well-known fact that just one-half of the sulphur in pyrites volatilizes very easily, and that the trouble is mainly to remove the remainder. It was thus apparent that the atmosphere surrounding the ore was of as much importance as the temperature. This led to the trial of a kiln to heat the ore by conduction through the brick. It was divided into ore-chambers, which were surrounded with flues, through which the products of combustion of gaseous fuel passed without coming in contact with the ore. We did not, however, succeed by this arrangement in getting more than a cherry-red heat, which, of course, was too low to accomplish anything. I had at this point become convinced – 1st. That gaseous fuel only would accomplish the work successfully. 2d. That since a high temperature was required to insure a rapid oxidation of the sulphur, and also to make the ore open and porous,

Jan 1, 1881

By W. J. Taylor

Some eight years ago I became interested to a considerable extent in one of the well-known deposits of sulphury iron ore in New Jersey, and, as a consequence, soon became interested in the subject of the elimination of the sulphur. On investigation I found that no department of pig-iron manufacture had received less attention than that of roasting or calcining ores for the elimination of sulphur, carbonic acid, water, etc. My first devices to eliminate sulphur mere naturally in the limekiln line, with solid fuel mixed with the ore, in the same way = common lime is burnt with small coal. While this process will succeed tolerably well for carbonates or calcareous ores, even though they contain sulphur as sulphates, I soon found it would not answer for our magnetites, which are very hard and dense, carrying from 2 to 5 per cent. of sulphur as pyrites, and become very sticky as soon as any attempt is made to roast them at a high heat. Some laboratory experiments developed the fact that the nearer the gaseous current surrounding the hot ore approached in composition pure atmospheric air the more rapidly the sulphur was oxidized, while in an atmosphere composed largely of carbonic oxide, carbonic acid, and nitrogen, it seemed almost impossible to reduce the sulphur more than one-half, which was a confirmation of the well-known Pact that just one-half of the sulphur in pyrites volatilizes very easily, and that the trouble is mainly to remove the remainder. It was thus apparent that the atmosphere surrounding the ore was of as much importance as the temperature..• This led to the trial of a kiln to heat the ore by conduction through the brick. It was divided into ore-chambers, which were surrounded with flues, through which the products of combustion of gaseoua fuel passed without coming in contact with the ore. We did not, however, succeed by this arrangement in getting more than a cherry-red heat, which, of course, was too low to accomplish anything. I had at this point become convinced— 1st. That gaseous fuel only would accomplish the work successfully. 2d. That since a high temperature was required to insure a rapid oxidation of the sulphur, and also to make the ore open and porous,

Jan 1, 1881

By Steve Kral

The domestic mining industry may be in the doldrums but there appears to be hope, at least from a regulatory standpoint. At least one US senator believes that the events of Sept. 11, 2001 made many citizens see the need for the country to become more self-reliant on its growing energy needs. Even before that tragic day, however, the Bush administration was making a strong push to open up more public land for energy and natural resource development.

Jan 1, 2002

By Damien Wagner

To reduce the emission of greenhouse gases by the steel industry, particularly for ironmaking, the production of DRI (Direct Reduced Iron) using hydrogen as the reducing gas instead of carbon monoxide is being considered. In this context, the reduction of pure hematite by hydrogen was studied at the laboratory scale, varying the experimental conditions and observing the rate and the course of the reaction. All the reduction experiments were performed in a thermobalance and supplementary characterization methods were used like scanning and transmission electron microscopy, X-ray diffraction, and Mössbauer spectrometry. The influence of rising temperature in the range 550-900°C is to accelerate the reaction; no slowing down was observed, contrary to some literature conclusions. A series of experiments consisted in interrupting the runs before complete conversion, thus enabling the characterization of partially reduced samples. Interpretation confirms the occurrence of three successive and rather separate reduction steps, through magnetite and wustite to iron, and illustrates a clear structural evolution of the samples. Finally, the influence of the sample type was revealed comparing a regular powder, a nanopowder and a sintered sample. The regular powder proved to be the most reactive despite its larger grain size, due to a more porous final structure.

Jan 1, 2006

By Tracy Jarrett

IN making gold alloys possessing certain physical properties, such important factors as corrosion resistance, color, hardness and melting points must be considered. All of these requirements depend directly upon the composition. It is possible to obtain the same colored gold alloy by using different compositions, but in doing so one may bring in some undesirable property. Other investigations have been carried on in this field by F. E. Carter,1 E. M. Wise,2 and L. Sterner-Rainer.3 To make a complete study of the effect of composition upon these properties, a systematic investigation was carried on involving a range of copper content from 85 to 65 per cent. Elements such as zinc, silver and nickel were used to replace in part the copper content. The gold content in all the alloys remained constant. Only 10-karat gold alloys of those containing 41.6 per cent gold will be considered. All of the alloys were made in an induction furnace. Each alloy was cast in a 3 by ¼ by ¾ -in. ingot. The method of obtaining the melting points and making color determinations was described in a previous paper.4 The Rockwell B scale with the ball and 100-kg. weight was used for all hardness determinations. The submersion of test pieces in concentrated nitric acid is one of the standard corrosion tests used in the optical trade. A test piece in the 50 per cent cold-worked condition, ½ by 2 by 0.043 in., is blanked from a strip using a. die, and the weight, loss per square inch of surface is taken as a means of comparison. The time for this test is 2 ½ hr. of actual sub-mersion of the test piece in the acid. This standard test was applied to all of the 10-karat gold alloys discussed in this paper. The tensile tests were made upon strips of the dimensions shown in Fig. 1. The test pieces after blanking were annealed at 1300° F. for 20 min. and cooled in air. The "drop of the beam" method was used in determining the yield point, as there was a very marked "drop" when the yield point was reached.

Jan 1, 1939

By Arthur G. Levy, Henry M. Howe

$1. Introduction.—Any given piece of metal is made up of a very great number of grains, usually microscopic, each of which is a perfect crystal save only in outward form, with cleavage planes of low cohesion, geometrically arranged quite as in fie familiar non-metallic crystals. The plastic deformation of such a mass occurs chiefly through the slipping of the crystalline blocks of which each grain is composed along these cleavage planes, causing steps called slip bands to form on a previously polished surface. In this slip, the metal immediately adjoining each of the cleavage planes along which the slip takes place is thought to pass extremely rapidly through a very mobile state into the amorphous state, in which it is harder than the remaining still crystalline metal between the amorphous layers thus formed along the slip planes. Whether the slip planes below the surface in ferrite can be detected by cutting, polishing, and etching sections is in dispute; but if this detection is possible at all, it is usually made impossible by heating the metal even gently. In short, the change along the slip planes does not persist through heating, and it is thought not to persist through time even at the room temperature, nearly all the amorphous metal recrystallizing. The network in Fig. 22 and the parallel N.-70-E. lines in Fig. 21 are slip bands in copper. When plastically deformed metal is heated, certain parts of certain grains may twin, that is, their component crystalline units may rotate into a position, or more exactly into an orientation, symmetrical with the initial orientation and hence with that of the remainder of that grain. As seen in a microsection of a metal the twinned areas usually have parallel sides. The existence of twinning is recognized in metallic sections either by the presence of such parallel-sided areas differing in brightness from the adjoining metal, especially when seen under oblique light after etching, or more surely by the zigzagging of the slip bands which form when twinned metal is itself deformed again. Hence the usual procedure in developing twins is to deform plastically so as to cause the twinning tendency; to heat so as to allow the tendency to assert

Jan 1, 1915

By Asmo Vartiainen

Background of Direct-to-Blister Flash Smelting is discussed; especially slag chemistry relating to copper solubility, slag fluidity and impurity behavior. In December, 2000, Mini-pilot Flash Smelting test runs were carried out at Outokumpu Research Oy to produce blister copper from Chilean concentrate and white metal and from a mixture of them. Tested slag types were iron silicate slag, ferrous calcium silicate slag and calcium ferrite slag with different slag compositions and different concentrate/white metal ratios. The test runs were successful. The fluxing effect of copper oxide was clearly demonstrated and ferrous calcium silicate slags remained fluid on a very wide liquidus range. The distribution of arsenic between ferrous calcium silicate slag and blister copper is much higher than between iron silicate slag and blister copper. This successful test run with the new slag type offer new possibilities in Direct-to-Blister process design.

Jan 1, 2003

Supplement to Proceedings Number 120, December, 1940. Drawings accompanying paper by T M Carey 'The Surface Plant at the Zinc Corporation Mine, Broken Hill'

Jan 1, 1940

By D. Argyriades, G. Derge, G. M. Pound

The electrical conductance of molten FeS was studied as a function of temperature and composition. It was found that stoi-chiometric FeS (36.5 pct S) shows a minimum specific conductance of 400 ohm-1 cm-1. For the Fe-rich melts the conductivity increases with increasing amount of iron to about 4850 ohm-1' cm-I at 26 pct sulfur. For the sulfur-rich melts the conductivity increases with increasing amount of sulfur until it reaches a maximum of 1300 ohm1 cm-1 for 37.4 pct S and decreases to 800 ohm-1 cm-1 for 38.3 pct sulfur. It is concluded that molten stoichiometric FeS is an intrinsic semiconductor in which the forbidden gap in the electron energy level diagram is quite narrow. Excess sulfur acts as an acceptor impurity in providing a component of positive hole conduction, while excess iron acts as a donor impurity to increase the n-type conduction. K. Bornemann and G. von Ftauschenplatl used a four-terminal, direct-current cell to study the electrical conductance of molten copper sulfide. Their measurements showed a high conductivity, of the order of 100 ohm-1 cm-" and a positive temperature coefficient. From electrolysis studies, Savelsberg2 concluded that molten Cu2S and molten FeS are electronic conductors. Further. he found that molten FeS has a high specific conductance, of the order of 1500 ohm-' cm-', and a small negative temperature coefficient. Pound, Derge, and Osuch,3 using an ac Kelvin circuit and a four-terminal cell, measured the specific conductance of molten Cu,S, molten FeS, and various mixtures. They found that the specific conductance of the solutions followed a roughly additive rule of mixtures. Yang, Pound, and Derge4 showed from electrolysis measurements that the mechanism of electrical conduction in molten Cu2S, FeS, and mattes is primarily electronic. Further, they observed that small additions of CuCl suppress the electronic conduction whereupon the system behaves ionically.

Jan 1, 1960

By A Purchase, C Stacey

The procedure for operation of deluge systems installed in road tunnels is still a subject of debate. The work undertaken in this paper assists in the understanding of the effects of deluge in road tunnels, and thus aims to help inform the development of operational policy. Analysis was undertaken using computational fluid dynamics (CFD) simulations to help explore the deluge interaction with the tunnel ventilation (including smoke exhaust) and combustion products. It was found, for the scenarios considered, that: Deluge activation can decrease the required critical velocity for smoke backlayering, and this in turn can improve the effectiveness of a smoke exhaust system.The dominant parameter for deluge effectiveness is the vehicle configuration, with shielded scenarios (such as a fire in the passenger compartment) showing almost no water reaching the fire. The water delivery rate (5 mm/min versus 10 mm/min) has only a minor effect on the smoke duct damper inlet temperatures and downstream smoke temperatures for small fires (5 MW). For large fires (50 MW) the higher water delivery rate gives slightly improved performance (lower damper inlet temperatures and lower downstream smoke temperatures). A one-dimensional transient tunnel model incorporating different operational and deluge response policies (ie deluge activation times) was developed. It was used to examine the influence of different operational parameters on life safety risk and tunnel damage risk. It was found, for the scenarios considered, that: The life safety risk for different scenarios has a sensitivity to deluge activation policy with the case of delaying the deluge activation for a long time giving the worst outcomes. However, the most critical parameters for life safety risk are operator response time (for all response activities, not just deluge) and the type of fire (configuration, size and growth rate).There is a reduction in risk of tunnel damage when the deluge is activated earlier, although the fire configuration (ie whether or not the seat of the fire can be reached) has the greatest influence on the outcome. It is concluded that the deluge activation policy should be case-dependent. It will depend on the tunnel, the safety systems provided, and most particularly the vehicles involved. The current policy in Australian tunnels is for manual activation of the deluge by the tunnel operator as soon as practical after the fire is confirmed (see also PIARC, 2007). This is a good policy. However, future work may be helpful to improve the skills of operators to make the best judgement as to when deluge should be activated, balancing all risks. This work could include research to further the understanding of deluge and fire in road tunnels, training programs and consideration of historical fire events.

Jan 1, 2008