The Homestake Pipe Conveyor

By Charles F. Langdon
Open Cut Mine Operations Superintendant
Homestake Mine
Lead, South Dakota

Presented at the
American Mining Congress Mining Convention '89
Materials Handling Session
September 20, 1989
San Francisco, California


Mining at Homestake began on a surface outcrop in the Open Cut in Lead, South Dakota, in 1876. Surface mining ceased soon after the turn of the century when operations progressed underground to follow the plunge of the ore body.

Early miners recovered the higher grade gold ores from a combination of timbered stopes and room-and-pillar mining methods. From 1876 to 1945, when gold mining was shut down by the War Powers Act, about 40 million tons of gold ore had been removed.

The underground mining resulted in ground subsidence. This subsidence caused a large portion of the main downtown section of Lead to be demolished and relocated further up Main Street west of the mining activity. The subsidence began in 1918, but the majority of the downtown area was demolished between 1928 and 1938. Cleared of downtown buildings and residences, the Open Cut became a Black Hills landmark on the edge of Lead.

When mining resumed after Wodd War 11 Homestake began sand back-filling the mined out stopes to curtail subsidence. As the miners followed the southerly plunge of the Homestake formation the Open Cut was left behind, remaining a page in recorded history in the development of a great mine. In the interim, the "Cut" had become a junkyard for the disposal of old refrigerators, stoves, car bodies, and mine timbers.


In 1980 Hornestake's interest in the Open Cut was rekindled when gold soared to nearly $850 per ounce and "non ore" became ore grade utilizing the lower costs of open-pit mining methods. Guided by old mine maps, a surface drilling program and underground drifting in the period from August 1981 to March 1983 identified a mineable reserve underneath the Open Cut.

Test Pit Ore Haul

During 1984 and 1985 a test pit was mined to confirm a portion of the reserve. The ore was crushed at the pit and hauled two miles over mine roads, city streets, and through a congested shop-yard to the mill. Hauling on city streets would only allow use of small 10-wheel trucks. Limited to a capacity of 16 tons, they had a gross weight of 46,500 lbs. There were complaints about engine and exhaust noise, dust, and mud tracks on city streets, as well as concern for safely operating eight to ten trucks in conflict with residential traffic. A day-shift operation was maintained to avoid disturbing the community at night. Homestake voluntarily shut the trucks down twice each day during school bus loading and unloading at a Main Street intersection. An overhead spray bar was employed to suppress dust blowing from the truck payloads. Streets were washed. three and four times a week to remove mud and tyre tracks. Winter storms required continuous sanding of steep city streets to maintain safe driving conditions. In the winter truck traffic compacted the ore stockpile into a mass of hard frozen rock. When dozed to the reclaim tunnel, frozen ore chunks tended to plug the feeders. The plugs also caused spills in the tunnel and production delays. Other problems bordered on the absurd: Homestake had to build a privacy fence for one resident who alleged that the truck drivers were "gawking" at her from the haul road 100 yards away when she sunbathed on her backyard deck.

The impacts of the truck haul confirmed Homestake's earlier predictions that another method of ore transport must be used.


Alternate Methods of Ore Transport System

The rocky steep terrain, residential properties, city streets, and highway confronted Homestake's engineers with problems selecting a transport route.

An aerial bucket tramway was considered but ruled out because of safety, high capital cost, and limited production capacity.

Preliminary engineering designs for a conventional conveyor system required a minimum of five sections. A disadvantage was that each section required its own drive system and transfer hopper with attendant, noise, dust, and maintenance problems.

The pipe conveyor concept offered a viable solution to the Open Cast's problems for ore transport over a circuitous route with minimum impact upon the community. The pipe shape allows the conveyor to be routed around horizontal and vertical curves. The pipe conveyors unknown in U. S. operations, had been in use since 1971 in South Africa, Australia, and the Orient.

The Pipe Conveyor Concept

A pipe conveyor loads and unloads like a conventional conveyor (see fig. 1 )After loading, the flat segment of the conveyor is folded into a pipe configuration by a series of steeply inclined troughing idlers and finally a series of six idlers in the form of a hexagon.

Engineering design determines a relationship between the belt width and the nominal inside diameter of the idler cluster to provide an overlap. This ovedap eliminates leakage of ore fines from inside the tube.

The pipe, or tubular shape, is maintained by clusters of six idlers, mounted in panels, at evenly spaced intervals. The belt is allowed to open over the head pulley and remains open as it passes through the drive and take- up systems. After the flat return section is through the take-up system, two tucking rollers insure the proper overlap before the belt enters the first cluster of idlers. The return cluster of idlers is mounted in the lower half of the panel below the load-carrying idlers.

The drive system consists of one or more motor-reduction gear assemblies, depending upon length and horsepower requirements.
Belt tensioning in a long conveyor is normally accomplished by hanging a pulley and counterweight assembly from a segment of the return belt snubbed between two directional pulleys. A variation of this tensioning system was developed for the Homestake conveyor and will be described in another section of this paper.

Homestake representatives inspected installations in South Africa and Australia. Positive reports prompted Homestake to request a proposal from Robins Engineers and Constructors. the U. S. licensee for the Japan Pipe Conveyor. The proposal was to design and build an operating system which would take -5/8 inch crushed rock at the mine site and transport it 6,300 ft. to Homestake's mill. This proposal when compared to other proposals indicated the cost for the pipe conveyor was higher than a conventional conveyor, but significantly less than an aerial tram. It was estimated that its lower operating and maintenance cost would more than offset its higher initial cost. Homestake engineers travelled to the Ohent to investigate several installations. The majority of the installations were less than 61 m (200 ft), but one installation visited was 457 m (1500 ft) long. It was handling coal at a cement plant in Taiwan.

Design Specifications

Convinced that the system was a viable alternative to a conventional conveyor, a more detailed proposal was prepared. The system had to deliver 318mt (350 tons) per hour and be able to start up fully loaded at -35 deg C (-30 deg F). The belt design was critical to withstand the site requirement for total conveyor length of 1920 m (6300 ft). The Bridgestone Company had not in 1986 manufactured a belt with steel fabric plies, so the system was designed using a belt with nylon plies. The belt would have to withstand the tensional load of it's own weight, the friction of 11,022 idlers and a contained load of 64 mt (70 short tons) under the worst possible conditions, -35 deg C (-30 deg F).

Preliminary designs indicated this system could be built achieving Homestake's requirement for an ideal ore transport system. The motor drives and transfer hoppers would be located at each end of the conveyor on Homestake property away from the residential areas. The pipe shape with overlap would assure dust and ore fines would be contained within the pipe. In addition, if the 6300-ft. length caused insurmountable operating problems, the belt would be cut into three lengths. This condition of the contract would be insured by a two million dollar bond.

The final cost comparison indicated the cost for the pipe conveyor to be about the same as a conventional conveyor. The latter included five transfer points and related dust suppression equipment.

The Japan Pipe Conveyor utilizes a belt 1350 mm (53 in.) wide fabricated by the Bridgestone Rubber Company. In its pipe shape the inside diameter is 305 = (12 in.). Belt edge overlap is 165 mm (6 in.). The belt is running at 2.67 m/s ( 525 ft/min ). It has an expected life of no less than five years at 453,600 mt (500,000 tons) per year. The equipment warranty provides for 100 percent replacement costs through the first year of operation, from 10 to 100 percent from the second through the fourth years) and a 10 percent replacement value for years five through ten. It is rated at 10,000 P.I.W. The pipe cross section is 25 percent full and is calculated to be carrying 70 tons of ore crushed to a -5/8 inch.

Route design changes reduced the final conveyor length to just under 6,000 feet. The belt length including take-up is 12,200 ft.


Details of the installation of the pipe conveyor are recorded in a report by Jerry Pontius, Homestake Maintenance Superintendant. For continuity of this paper, highlights of the installation are reviewed in the following paragraphs.

The conveyor was installed during the last six months of 1986 during very mild autumn and winter weather. Frection was by The Industrial Company (T.I.C.) under the direction of Robins Engineers and Constructors. Survey, site preparation, and the pouring of concrete support piers were completed in July and August. Installation of factory-fabricated support modules commenced in September. Adjacent modules spanned the piers. The modules were 7.6 m (2 5 ft) or 10.4 m (34 ft). The shorter length was used for installations on horizontal and vertical curves; the longer length was used on straight-line installations. Horizontal curves were designed with a minimum radius of 500ft. The steepest slope is a downhill segment at 1711(30%) The entire length of the conveyor is enclosed within framed modules. The modules are covered with corrugated siding for protection from ice and snow. An exception is a segment enclosed in a 2 112-m (8-ft) diameter culvert passing under a haul road to a waste dump. The module enclosure is 71 cm. (28 in.) wide by 142 cm (56 in.) high. Removable door panels provide access to each idler cluster, which are mounted in framed steel panels at intervals of 1.8 m (6 ft).

Approximately 11,022 idlers4 displays: each 102 mm (4 in.) in diameter by 182 mm (7 31'6 in.) long, were factory installed, and the critical end gap adjusted from 3 to 5 mm after start-up of operation. These idlers use factory-lubricated, sealed bearings. Modules for horizontal and vertical curves were scheduled for a specific site location. The idler support panels for these modules were oriented within the module frame at right angles to the longitudinal axis of the conveyor tube.

The belt was installed in 9 segments, each about 418 m (1,3725 ft) long. Each segment was vulcanised to the preceding segment and the connected segments pulled through the forming idlers with a truck-mounted winch operating from alongside the conveyor.

Belt Commissioning

A team of engineers under the direction of Mr. Shinsuke Hashimoto, the inventor, arrived and began the six-week commissioning period on Decernber 22, 1986. This process involved installation of wedges behind the idler brackets to induce rotation of the belt where needed to maintain the belt overlap in the upper half of the quadrant. Twisting and fold-over at the head pulley were frequent. Success had not been achieved by Christmas when the working crew from T.1 C. went home for the holidays.

After the new year the Japanese team returned with a new leader, Mr. M. Sankae, Chief Engineer for the Japan Pipe Conveyor Company. This team removed many of the shims and relocated others installed by the first team. The belt required a period of "breaking in" or "taming." During the period, rotation of the overlap stabilised and the surface of the belt appeared to take on a grey color. Other adjustments included elevating the head end pulley about 165=(6 inches) to improve the geometry for the pulley to open the pipe. Modifications were also made to the feed hopper and level control gate at the feed end. A fluid coupling was installed to replace the solid coupling on the tail drive.

Final acceptance by Homestake was based on a continuous running of the belt for 72 hours maintaining the overlap in either of the upper two quadrants. The belt was scheduled to run loaded on day shift only. Final acceptance was achieved on February 5, 1987.

Route Description

After loading at the Open Cut, the conveyor climbs like a roller coaster vertically 47 m (154 ft) at 7.25deg (13%) to the 5360 bench on the east side of the Open Cut. It runs downhill dropping a vertical distance of 78 m (257 ft) before crossing the highway on an elevated trestle. South of the highway the conveyor begins a long, looping, horizontal curve and uphill ascent of 41 m (133 ft). The remaining section of the conveyor to the discharge end follows the contour before climbing to discharge into a series of chutes in a reclaim tower. From the feed end to the discharge, the conveyor route is in the shape of a question mark. The discharge elevation is only 32 ft above the feed end elevation. At this point ore can be transported directly to the mill fine-ore storage bins or diverted via a reversing conveyor and stacker to an outside surge pile.

The steepest sections are on both sides of the roadway trestle crossings. The steepest grades for the loaded belt are downhill at 17deg and uphill at 13.5deg.

With the exception of the elevated section across the five roadways and the elevated discharge end, the conveyor follows the terrain. The top of the enclosure is approximately 1.8 m (6 ft) above the ground. The elevated sections over the roadways are supported by, prestressed concrete T-beams and A-frame steel bents.

Drive and Counterweight System

The three head end motors are energized sequentially with no load to minimize line overload. The 200-hp, TFC, fan-cooled, Westinghouse motors drive the conveyor through Falk-Sime 177HC scoop control drives and Hansen model RDF31 -AND reduction transmissions. The scoop drives take the belt from start-up to 160 m (525 ft) per minute in 45 seconds. The start-up and acceleration to operating speed is unusually smooth and shock free.

The three head pulley motors start the belt moving. Approximately 45 seconds after the head end drives initiate motion and the belt stretch has been taken up by the trolley and counterweight assembly, the tensioned top section of the belt begins to rotate the tail pulley. A proximity switch sensing rotation of the tail pulley energizes the 200-hp tail pulley motor through a fluid coupling and a model RDF31-AND Hansen transmission. The reduction ratio is 32.2:1. If maximum operating speed is not achieved in 90 seconds the programmable Alien Bradley model 230 operating computer de-energizes the motor circuits. This procedure assures all four motors are operating and no unusual friction forces are restricting a normal start-up.

A segment of the belt downstream from the head pulley passes around both the primary and secondary rubber-lagged drive pulleys. Snub pulleys of the same size insure a 1800 wrap on both drive pulleys. Power is transmitted to the drive pulleys from the motor/scoop-drive power-train systems located on concrete pedestals. These pedestals position the drive systems about 1.2 m (4 ft) above ground level, a convenient height for inspection and maintenance.

The take-up system is located in the return belt section between the head end drive and the first panel of pipe-forming idlers. The extraordinary length of take-up required, by a belt 6300 ft long, required the take-up pulley to be mounted horizontally on a track-mounted trolley. The trolley is connected through a system of directional sheaves to a 33,000 lb. counterweight hung from a tower 54 ft high. At start-up the top or loaded section of the belt is tensioned and stretched by the head end drives. The excess belt slack is pulled through the drive system and over the trolley sheave by the counter- weight until the return segment can catch up with the loaded segment. The elasticity of the 6,000-ft-long conveyor belt is noted in the travel distance of the trolley on start-up. During the warm summer months this stretch has been measured as much as 1 0 m (33 ft) or 0.6 percent of the belt length between the head end and tail pulleys. The belt stretch at start-up during the cold winter months has been measured at less than half this distance. Before passing into the first pipe-forming panel for the return trip, two "tucking" rollers mounted at slightly different elevations insure that the belt edges do not switch positions in the overlap. These rollers, two at each end of the conveyor, are cantilevered from a single end-mounted shaft. The opposite end of the roller is rounded or "bull nosed". The belt overlap is about 165 mm (6 in.). The 1,350-mm-wide (53-in.) belt is extremely pliable and is easily idled and maintained by the forming idlers in a 12 inch diameter tube with a closely fitted overlap.

The Bridgestone Belt

The belt is fabricated with three nylon plies. The three plies extend to within 13 mm (0.5 in.) of each edge. The top and bottom covers are 5mm (3/16 in.) and 3 mm (1/8 in.), respectively. The belt thickness is 12 mm (7/16 in.). The close proximity of the plies to one another and a composition that is 45 percent natural rubber undoubtedly contribute to the belt's pliability. The rubber compounding provides an extremely hard, abrasion-resistant surface.


The Open Cut's responsibility is to "fill the mill", which is making up the difference between the mill capacity and what the underground mine can provide. A targeted combined mining cost of the underground and Open Cut currently dictates that the majority of the mill feed be underground ore (80 percent) and the balance (20 percent) be Open Cut ore.

Annual Open Cut Production
Milling Capacity 2,268,000 mt (2,500,000 STPY)
Underground Mine  1,769,000 mt (1,950,000 STPY)
Open Cut Mine 499,000 mt (550,000 STPY)

This tonnage is crushed and conveyed at the rate of approximately 9,500 mt (1 0,500 short tons) per week. The pipe conveyor is required to operate only 30 to 32 hours per week. The operators conduct routine maintenance and cleanup on the crushing and conveying equipment or work in other mine operations when not actually operating the pipe conveyor.

The system was designed to convey a maximum annual tonnage of 635,000 mt (700,000 tons) during two years of peak Open Cut production in 1993 and 1994. This tonnage will average 12,250 mt (13,500 short tons) per week operating about 39 to 40 hours per week. The greater workload occurs on weekends when shaft inspections and maintenance limit the amount of ore which can be hoisted from underground. This year the Open Cut mill requirement is only 522,300 tons or 21 percent of the total mill budget. This quantity can easily be conveyed averaging 28 to 30 operating hours per week. Effective utilization (year-to-date as of 7/31/89) is only 38 percent.

Mines do not always run smoothly and the Homestake is no exception. When production from the underground is reduced or curtailed for any number of reasons from shaft problems to unavailability of ore, the Open Cut is called upon to make up the difference. The pipe conveyor has been able to maintain the required feed allowing the mill to operate at 100 percent of capacity. Another beneficial result has been a reduction in total milling costs of about 20 percent, or approximately $1.50 per ton.

When called upon to perform, the conveyor has done so admirably. During an eight-day period in June 1989 when production was curtailed from the underground because of a mine fire, the Open Cut conveyed just under 36,300 mt (40,000 dry short tons) averaging 4,540 mt (5,000 tons) per day. A record production of 7,521 wet tons was conveyed through the pipe in one 24-hour day during this period. June 1989 was a record month for the pipe conveyor. A total of 70,756 mt (77,994 short tons) were conveyed. Availability for June was 100 percent.

The pipe conveyor has transported 1,248,000 mt (1,376,000 short tons) during 30 months of operation ending July 31, 1989.


Crushed ore is reclaimed from a crushed ore stockpile through feeders located in a 3 m (10 ft) diameter underground reclaim tunnel. The feeders discharge onto a belt which discharges into a lump breaker and over a vibrating grizzly to insure frozen chunks do not enter and plug the pipe conveyor feed hopper.

During crushing operations a travelling stacker replenishes the stockpile. When the crusher is not operating and the live cone sections in the stockpile over the feeders have been pulled, a D7H dozes crushed ore to the feeders. This dozer operator, a member of a two-man team which runs the conveyor, also inspects all systems before start-up. Start-up can only be achieved at the-Open Cut from either the Alien-Bradley programmable controller or a remote station in the reclaim tunnel. The second member of the team has assigned duties at the discharge end of the pipe at the mill. He performs a pre-shift inspection and operates and monitors nine conventional conveyors reclaiming the ore from the pipe conveyor. All conveyors are equipped with electrical safety interlocks.

The conveyor is manned to operate 2 eight-hour shifts per day, 7 days per week, 365 days per year. Three two-man crews are scheduled, with one crew always off duty.

Power Requirements

During the winter months from mid December to mid March when the daily minimum temperature drops below 32' C, all four 200-hp motors are utilized to avoid excessive horsepower loads on start-up. During the remainder of the year when warmer temperatures improve the pliability of the belt and warm the lubricant in the idler bearings, the belt is successfully started and operated utilizing only three motors, two at the head end and motor no. 4 at the tail end. Meter readings during the warm weather months indicate a combined load of 188 hp at the head end and 76 hp on the tail pulley drive motor.


Early inspection procedures were limited to a walk-along over the length of the operating conveyor listening for the telltale sound of a noisy ball bearing. This method did not address a specific type of failure: A worn bearing assembly during its early phase of failure develops a low level sound different from that of a bearing in good condition. Eventually the bearing becomes noisier and disintegrates. The idler housing drops down onto the idler shaft. The moving belt continues to drive the idler, and the idler hm opposite the failed bearing makes contact and wears through the support bracket. Failure of the bracket exposes a knife edge to the moving belt (see fig. 5). Experience has indicated the bearing failure must be corrected during the early phase of failure. If operating time is limited to about 100 hours, about three weeks, the potentially damaging failure condition can be avoided.

Experience identified another maintenance problem. The air gap between the rotating edges of idlers must be maintained between 3 and 5 mm (1/8 in - 3/16 in.). The belt rotates (1) between loaded and unloaded segments, (2) during subfreezing or hot weather conditions, or (3) because of any influence resulting from tension. If operators find the-belt acting erratically over a prolonged period of time and the behaviour cannot be explained due to temperature or intermittent loading, they will investigate other conditions which cause tension and belt twisting. An unstable belt will rotate its overlapping edge into a gap between adjacent rollers if the gap is in excess of the belt thickness of 12 mm (7/16 in.) On one occasion inspectors witnessed the belt running between adjacent rollers for a distance of 30 ft. The belt surface is also exposed to scuffing by the idler shaft ends and retainer cap screws when running between idler brackets.

Spillage has occurred on very rare occasions and has been limited to a three-to seven-inch accumulation on the floor of the conveyor enclosure over several dozen feet.

These types of maintenance problems have evolved into an inspection procedure which necessitates removing the doors to inspect the idlers for wear and acceptable spacing. Nine months after start-up worn idlers were recognised and replaced. During the next 21 months of operation, idlers were replaced at the rate of 31 per month or 19 per 100 hours of operation. Two employees can change out 19 idlers with brackets in 4 hours.

The conveyor belt has to date required only two cold patches for a pair of cuts about 16 cm (40 in.) and 6 cm (44 in.) long. The Pang cold-patch system used for the repair was developed in West Germany.

Equipment Warranty

The contract with Robins provided for an equipment warranty which replaces the conveyor belting and drive systems up to 10 years or when five million tons of gold have been transported by the system, whichever occurs first. This guarantee covers normal wear and tear under the design parameters that no more than 350 STPH are conveyed and the minimum temperature that the pipe conveyor will operate is no less than - 35deg C (- 30deg F).

Year Item Purchase Price Conditions
1 Replacement belt or portion of replaced belt 100% --------
2-4 Same 100%-10% Applicable percentage
5-10 Same 10% --------

The "non-conforming" components of the drive systems identified as the motors, scoop drives, Hansen reducer, and fluid coupling will be replaced at no cost to the owner when returned to the factory.


After commissioning, the conveyor was turned over to Homestake operating personnel on February 5, 1987. An unusually mild winter and spring did not allow the operating crew an opportunity to test the stability of the conveyor belt under winter operating conditions.

On March 25th the first incident of the belt failing to open at the head pulley occurred. Operating personnel appropriately named the occurrence a head pulley fold-over. The system shuts itself down when an alignment switch is topped or if the head speed drops below a specific rate. We were uncertain as to the cause of this fold-over but noticed it occurred with only a partial load on the belt, resulting in what was later identified as a tensioning of the belt and subsequent rotation of the overlap. When the overlap or "seam" rotated out of either of the upper two quadrants, the belt resisted the head pulley's effort to open it.

The second incident happened on April 12th when the take-up carriage jumped the trolley rails. The cause of this mishap, a gap in the rails and too close a tolerance on carriage to rail retainer brackets, was corrected. It is unknown if belt tensioning may also have contributed to the mishap.
An unrelated incident to the head pulley fold-over was the discovery of damaged idler brackets and panels on the trestle crossing over the highway. Neither Robins nor Homestake engineers could identify the cause of the damage. The idler brackets were either deformed or torn from their mounting panels. The pipe circumference appeared to have expanded due to an overload in the pipe occurring along an approximate 100-ft-long segment. This type of damage has never occurred upstream in the interim distance of several thousand feet to the loading point. Because the damaged section is located at the lower end of a 12deg downhill vertical curve, it has been postulated that fine dry ore in the pipe slides downhill from a loaded into an unloaded segment when the belt is stopped.

In mid July the counterweight rope sheave bearings failed. Modifications were made in the replacement assemblies to correct the problem.

During the summer and autumn a third problem was developing, which was more annoying than damaging. Operating personnel were experiencing an increasing number of PC (programmable controller) shutdowns, or "crashes". Located in the crusher building control room, the Alien Bradley PC initiates all the operating functions and interlocks all the safety controls. One operator from a single control station can start up, operate, and monitor not only the pipe conveyor, but also the tertiary crushing system. When the PC crashed, all crushing and conveying systems were simultaneously shut down. Homestake electricians installed isolation devices on head and tail end power racks believing electrical interference was affecting the PC operation. Another probable cause was the discovery of an unshielded data highway cable in the same conduit with the 440-volt power supply to the 200-hp tail pulley motor. Also in the conduit were 110-volt motor control circuits to the reclaim system conveyor drives. Power isolation devices were installed, but the problem was finally identified when the data highway cable was temporarily isolated from the power and control circuits.

On September 13th a third fold-over occurred that was attributed to a twist or rotation in the belt resulting from a partial or light load approaching the head pulley.

The fourth and fifth fold-over occurred in October and late November. Also, the system had experienced just as many incidents of damaged roller brackets and panels on the trestle as head pulley fold-over by early December.

Engineers were no closer to a solution to either apparently unrelated problem. During this time the operating crews did improve their procedure for efficiently pulling up and hanging the counterweight to gain slack in the belt. This slack was then winched over the head pulley to allow the manual folding of the belt to a normal flat position over the head pulley.

Enough incidents had happened to associate the fold-over with interruptions of feed onto the pipe conveyor. Observations had confirmed an 80deg-90deg clockwise rotation when an unloaded segment followed a loaded segment. The rotation was sufficient to move the overlap out of the upper 180deg quadrant in the 12 o'clock position into the lower quadrant. The head pulley could not open the belt and a clockwise head pulley fold-over would occur. It was assumed that differential tensioning had to be occurring between the loaded, lightly loaded, or empty segments resulting in a twist or rotation of the belt. Operations personnel realized that interruptions of feed onto the pipe conveyor had to be minimized to maintain belt alignment. These interruptions were caused by: (1) an overtime interval exceeding the time programmed into the controller at start-up to achieve pipe conveyor operating speed before feed could be introduced, (2) metal detection equipment shutting down the reclaim system feeding the pipe conveyor, (3) slipping and shutdown of the conventional conveyor feeding the pipe conveyor caused by a frost-covered dove pulley, and (4) running the entire load off the pipe conveyor at the end of each shift when the temperature was below freezing. Because feeding the mill from the Open Cut was not a continuous operation, it was standard practice to empty the belt after each operating shift during the winter months. Homestake did not know, and therefore dreaded, the consequences of an ore freeze-up within the pipe for the length of the conveyor.

During Christmas week the night-time temperatures dropped from the mid teens to below zero. At start-up on December 22nd the belt folded over. This fold-over was in the opposite direction of the normal counter-clockwise fold. Another variable, subzero temperature, was added to the list of factors causing instability of the pipe. What followed was "hell week" when there occurred multiple foldings of the belt every day from Christmas Day through December 31 st. During this period the weather was bitterly cold. The week was climaxed by forgetting to replace the load level control gate at the feed end after completing welding repairs to a feed hopper liner. An overload was allowed to enter the belt, physically expanding the circumference of the belt larger than the opening of the forming idlers. The resulting impact of the belt attempting to get through the first dozen forming panels tore 32 idler brackets from the panels. After this repair was completed the belt folded over the head pulley; and when the 16-ton counterweight was lifted to obtain the necessary slack, the wire rope clips decided it was their turn to fail. The counterweight dropped 15 ft to the base of its enclosure with no injury to personnel or equipment. Records show one or more fold-over were unfolded on each of 12 days during 1987. The operating crew began the new year with another fold-over on January 1st. On December 29, 1987, the temporarily re-routed data highway cable was reburied in its own ditch away from the motor supply and control cables. This permanent re-routing solved the PC crashes.

There was no relief from the bitterly cold weather or the fold-over problem during the first week in January. On January 3rd, while walking the belt after a fold-over, a 2.25 revolution clockwise twist was mapped followed by a 2 counter-clockwise twist. The latter twist occurred over a length 550m (1800 ft) or six lengths of a football field. It was impractical to consider manually untwisting this belt. A cantilevered tucking roller was installed at the head pulley to hold down the belt while the system was started up, loaded, and the twist "run out". This "hold down" roller was successful in restraining the belt from a fold-over. The fold-over had always been counter-clockwise up to this time, so the hold-down roller had been installed on the right-hand side to restrain a roll in this direction.

On January 7th a feed hopper that was plugged with frozen ore chunks overflowed onto one side of the tail pulley. The belt continued to run while the spilled fines continued to run onto the belt, rolling it over on the tail pulley 180deg. Ore that was still able to feed down through the hopper was now being loaded on the opposite or "clean side" of the belt. A complex S-shaped internal fold occurred between the two loaded sections. The fold was removed by hand only after removing 270 roller brackets. This was fold-over number 13 and it happened when the wind-chill factor in Lead was -46 deg C.

Murphy's Law was really at work. The belt was now loaded on both sides for several hundreds of feet with a 180deg twist in the middle. To correct the problem, crews manually folded the belt over the tail pulley 180deg in the opposite direction. The pipe conveyor was started up and loaded in the usual manner, and the twisted sections discharged their loads in a normal over the head pulley and unwound behind the hold-down roller like a rubber band.

Having returned to normal operations, procedures were employed to minimise interruptions in loading. A simple test was performed to confirm the belt did not have to be "run out" and emptied after each use. The Open-Cut crushed ore contains only 5 percent moisture, About 23 kg (501b.) of ore were placed in a container for 36 hours in temperatures at -29deg C. The frozen ore was chunky but did pour out of the container. With that knowledge we began leaving the belt loaded after shutdown for extended periods of time when the temperature was below zero.

Secondly, a large permanent magnet was installed on a conventional flat conveyor feeding the pipe conveyor to remove century-old scrap iron, nails, and spikes that were tripping the metal detector and shutting down the conveyor feeding the pipe. The permanent magnet significantly reduced not only the interruptions in feed to the pipe conveyor, but also the annoying "search and discard" procedure in a building away from the regular operators duties.

The next operational procedure change to maintain a load on the pipe conveyor was to reprogram the programmable controller to stop the pipe conveyor when the conveyor feeding it was stopped for a metal detector trip.

Shim Removal - Phase one

Engineers from Robins had been very co-operative with assisting Homestake in correcting operational problems. The most critical problem was the,bead pulley belt fold-over because they interrupted production and were labour intensive to remove. These fold overs did not appear to be causing any physical damage to the belt or systems.

After a site visit in mid January 1988, Robins Engineers recommended partial removal of the approximate 400 shims. They believed the shims exerted too much rotational effect to keep the seam upright over the length of the conveyor when extreme low temperature conditions were already causing tension and resultant rotational forces in the belt. If the shims were removed the return side or empty section of the belt would remain in its normal position with the overlap on the bottom. This is attributed to the greater weight of the pipe cross section at the overlap. The seam on the loaded or ore-carrying side should remain in its normal position on top as it leaves the loading area at the tail pulley. The weight of the ore on the bottom of the belt was expected to compensate for the tendency of the belt overlap to roll over to the bottom.

Robins made a request to the Japan Pipe, Conveyor Company for review and comment on installation of the hold-down roller and shim removal. They replied that they wanted more time to evaluate the recommendation. Robins went ahead with plans for shim removal. Shim removal, supervised by Robins personnel, commenced in early February with predicted rotation of the seam to the desirable locations.

Frozen Ore Accumulation on Conveyor Belt

A lesser problem, which continued to be troublesome during the winter months, was the accumulation and re- freezing of ore fines on the snub pulleys at the head end drive system. The pre-cleaner and Martin Durt Hawg scrapers on the head pulley were only partially successful in cleaning the belt. This accumulation increased the circumference of the snub pulley. The head speed proximity switch was reading a false "slower than allowable" rotating speed of the snub pulley. The result was frequent shutdowns and aborted start-ups. Both snub pulleys had to be thawed out and cleaned before and during shift operations. In addition, the build-up was unevenly distributed across the face of the snub pulleys causing a misalignment of the belt to one side through the drive system. During the 1987-88 winter season the use of Ice Free Conveyor 9, a spray-on antifreeze solution, was very successful in keeping the belt clean after the belt self-cleaned during dry weather periods. The compound did not work as successfully during the 1988-89 winter season. Homestake will continue to try various mechanical scrapers this winter to control the problem.

Shim Removal - Phase 11

Encouraged by the success of removing shims from the north half of the belt in February, Robins personnel returned to supervise the removal of the balance of the shims in April. Shim removal was completed on April 28th. After shim removal, belt stability was noticeably improved and fold-overs were virtually eliminated during the remainder of 1988.

Other Modifications

A magnetic head pulley installed on a crushing plant feed conveyor made a significant improvement in the removal of old mine nails and spikes. In late June, a Ding's electromagnet was installed upstream from the permanent magnet in the conveyor stream feeding the pipe conveyor. This magnet is collecting the balance of the old iron trash that was tripping the metal detector and shutting down the pipe conveyor. Unscheduled shutdowns responsible to iron trash have been 99 percent eliminated.

Shim Removal - Success

Shim removal appeared to be successful in maintaining a stable belt condition during the summer months. Operating personnel were hesitant to claim total success until the conveyor had operated successfully through a normal Black Hills winter. The opportunity arrived early last February when temperature plummeted from the mid 60's F on January 30th to reach a low of -33deg F on February 3rd. The conveyor was started fully loaded on February 2nd without mishap at -30deg C. Prior to start-up, the conveyor had been shut down with a full load on the belt for 40 hours in temperatures ranging from -23deg C to -33deg C. Operating personnel were jubilant. The conveyor continued to run uneventfully throughout February, the coldest month in recent recorded history.


The capital cost for the design and installation of the 6,000-ft long conveyor was $3,8 million, including site preparation, walkway, and safety guards. The total cost included $550,000 for the conveyor belt.

The Homestake cost accounting system combines operating costs (less depreciation) for the pipe conveyor and reclaim systems at the Open Cut and South Mill. This includes maintenance costs for a reclaim dozer, six feeders, nine conventional conveyors, two lump breakers, and a vibrating grizzly. Cash operating costs for the system were highest in January 1988 at $1.75 per ton due to a series of pipe conveyor fold-overs. After shim removal and other modifications, cash costs had dropped to a low of $0.39 per ton in February 1989. The year-to-date (July 31, 1989) average cash cost was $0.53 per ton. Including depreciation the total cost is $1.15 per ton. As utilization increases during the early 1990's these costs are expected to fall dramatically.


During the 30 months since start-up the pipe conveyor has transported 1,248,000 mt (1,376,000 short tons) of minus 16 mm (5/8 in.) feed to the Homestake mill.

Correction of operating problems during 1987 has improved availability from the mid 70's to 94 percent.

The system achieved a production record in June of 1 989 when 70,762 rnt (78,000 short tons) were conveyed. June availability was 100 percent. A new 24-hour record of 6,914 mt (7,621 short tons) conveyed was also set in June.

Elimination of intermediate transfer points has resulted in a quiet, dust-free, low-maintenance-cost operation compatible with a need to minimize the environmental impacts of mining in an urban area. The pipe conveyor has been the better choice for an efficient transport system operating over the rugged terrain of Laad, South Dakota.

A product from Japan, this first pipe conveyor in the United States and the longest in the world would not have been successful without the engineering by Robins Engineers, the erection by T.I.C., and the perseverance of Homestake's Open Cut operating personnel.


Toms, D., W. Stone, and G. Motchenbacker. The Gold Belt Cities - Lead and Deadwood. Photographic history of the gold mining communities of Lead and Deadwood, S. D. Published 1988 by Gold Unlimited.

Loving, G. A. and D. G. MeDowall. "The Open Cut Mine at Homestake." Paper presented at 1986 A.MC convention in Las Vegas, Nevada.

Pontius, Jerry. "Selection, installation, and Operation of the First Pipe Conveyor in the United States." Paper presented at the 3rd Western Regional Conference on Precious Metals, Coal and Environment. September,1987.

Idlers manufactured by Precismeca Company, Alabaster, Alabama.

Hiromi Kikuchi, Assistant Manager, Conveyor Belt Engineering, Bridgestone Corporation, May 1987.

Belt sections vulcanized by Sonic Systems, Denver, Colorado and Gillette, Wyoming.

  1987 1988 1989
Tons conveyed per month 38,394 42,603 57,726
Tons per operating hour 263 298 311
Operating shifts per month 42 39 45
Tons per shift 906 1,107 1,279
Average operating hours per shift 3.8 3.7 4.1
Total hours per month 466 503 503
Operating hours per month 159 143 186
Standby hours per month 271 321 304
Repair hours per month 48 39 12
Mechanical 9 5 7
Electrical 4 6 3
Belt 24 20 1
Other 11 8 1
Physical Availability 90% 92% 98%
Mechanical Availability 75% 79% 94%
Utilization 35% 31% 38%
Cash cost per ton $0.53 $0.62 $0.73
Total cost per ton $1.36 $1.32 $1.15