Backfill distribution system design
Traditionally, backfill has been produced from one source—the milled waste ore from the mineral extraction plant. This fine material has been, and still is, the preferred source of backfill material although milling technology has resulted in a finer tailings product and backfilling has moved from cyclone classified tailings to full plant tailings and higher solids concentrations as mining activity has resulted in larger cavities underground. These larger cavities, together with the use of non-draining backfills, has also increased the use of binder additions, which in turn, has prompted higher solids concentrations to reduce the binder requirement.
Backfill is delivered underground, through a system of pipes, under the action of pumps or gravity, or a combination of the two.
In Canada, paste backfill—where the solids concentration is maximized based on particle size distribution—is generally produced by blending the available aggregates to give the most desirable set of support and transport characteristics, or densifying metallurgical tailings to produce the required product. A general rule of thumb here is that the paste backfill should contain at least 15 to 20 mass percent solids smaller than 20 microns.
Since 1985, a great deal of research and practical experience in South Africa has gone into developing backfill distribution systems that reached to depths of 2500m and today reach 3500m below surface. The result of this experience is an extensive understanding of the required transport characteristics for backfilling operations, which has led to the development of two backfill distribution system configurations, both in use, successfully, on South African gold mines:
- Dedicated (uninterrupted) system
- Pressure break or underground dam system
The dedicated system
The dedicated system uses a single pipe range from the surface to the area underground to be filled.
Fluor Engineers Inc, 1986, in their research report, selected the dedicated distribution system for The Chamber of Mines as the most suitable backfill distribution method. Their system involved in the order of 32 dedicated pipelines from the surface and was selected because of its lack of valves in the shaft, decreasing the possibility of blockages, its flexibility and possibilities for back-up. This is the basic system that was functional in the South African gold mining industry at the time.
The system adopted for transporting the backfill underground through the pipeline was a gravity-feed system. This means that the storage tank on surface was connected into the underground distribution pipe ranges and the backfill was allowed to feed through the pipes under gravity. After the area underground was filled with backfill, the pipeline was flushed with clear water. Examples of dedicated ranges can be found on mines around the world.
The advantages of this system (Thomas et al, 1979) are:
- Surface activities are not disrupted.
- Pumping is (usually) not necessary.
- The distribution system is accessible from the different levels, for maintenance and routine inspection.
The disadvantages, again after Thomas et al, are:
- Clutter and activities in the shaft are increased.
- Pipeline failure results in shaft downtime.
- Pressures in the pipelines are high.
- Each level requires a dedicated pipeline.
- Pipelines clutter the respective levels and failure results in operational downtime of that level.
- Pumps may be necessary if the horizontal leg is long.
- There is a tendency for the solids to settle out in the horizontal leg.
The dedicated backfill distribution method is can be use in two general designs: free-fall and full flow backfill systems.
The initial design of backfill distribution systems involved both free fall and full flow dedicated systems.
Free fall systems
Free fall pertains to a transport condition existing in the vertical column in the shaft, where the solids fall in air to the backfill free-fall zone interface. The position of this interface in the vertical column, and the pressure head that it develops due to its height, is dependent on the pipeline pressure drop, or frictional losses, due to the flow rate of the specific backfill within the pipe. Thus if the air-backfill interface height increases due to an increased backfill delivery rate an increase in the flow rate in the full flow section can be expected, and this results in an increase in the pipe frictional losses resulting in the increased interface height. Guest et al (1986) state that because of this automatic pressure-regulating feature of the free-fall system, accurate predictions of pipeline frictional losses are not required. If any changes are made to the system, the interface height can be regulated until the required pressure head is achieved to drive the backfill at the required flow rate. They further suggested that the predictions of pressure gradients should be conservatively high in order to ensure the safe pressure rating of pipes.
The free fall system is shown in Figure 1.
Figure 1: Free fall
The delivery of backfill from the surface storage to the pipeline is through a flooded surge cone with a restriction orifice in the pipeline to regulate the flow rate. These orifices will require regular replacement to maintain a balanced system: in Canada a shut-off valve of some description is used.
The main advantage of this method of backfilling is that high pressures can be avoided in the main shaft. However, the disadvantages of free fall are numerous and tend to result in catastrophic failure:
- Inlet static pressures are below atmospheric, and therefore, air is sucked in through pipe joints and pipe linings can be pulled off, Thomas et al 1979.
- Excessive pipe wear occurs due to the high velocities of the particles in free-fall.
- Pipe bursting failure occurs due to cyclic impact loading at the air-backfill interface Lazarus and Paterson (1988), Thomas et al (1979).
Thomas et al (1979) gives a comprehensive study of the free fall system, its operation and pressures, Figueiredo (1987) and Guest et al (1986).
Full flow systems
Full flow systems are generally used in backfilling operations today: either as a dedicated system or a pressure break system. A full flow system is described by Thomas et al (1979) as one in which, “the system operates at the velocity at which the system pressure losses, due to pipe friction, bends, entry and exit orifices, are just balanced by the gravity head due to the difference in elevation between the discharge point and the free surface of the fill in the agitator tank.” A full flow system is shown in Figure 2.
Figure 2: Full flow
For such systems, no air-backfill interface exists and the backfill fills the system from the surface reservoir to the exit at the paddock. This balance has typically been achieved using sections of pipe in the backfill column that have a smaller bore than the overall column, resulting in increased frictional losses due to the increased backfill velocity.
The advantages of a full flow system are:
- Increased pipe lifetime, and therefore, reduced downtime and shaft time for pipe repairs.
- A controlled system.
The main disadvantage of a full flow system is the high static pressures that can be experienced if the pipeline blocks. Initially, this was the main reason for choosing free-fall backfill systems, which have far lower pipe pressures under static conditions; however, many pipelines were not designed with an overflow in the pipeline such that if the line blocked it would fill to the surface resulting in the full static head being experienced anyway. Considering the poor performance of free fall systems, the expense in worn and blocked piping, as well as boreholes, and the resultant down time incurred, the design of free fall systems will not be considered.
Pressure break system
The pressure break system was developed and introduced into South African gold mining for two reasons (Van Der Walt 1988):
- To reduce the high pipeline pressures in the shaft, and
- To reduce the number of backfill pipe ranges in the shaft.
A backfill storage dam is built underground on a level sufficiently elevated above the mining stopes to enable backfilling from the dam to the stopes to be carried out under gravity. The dam must be built in an area where there is sufficient space for a large reservoir of backfill and the associated operational controls. This availability of a dam area should be balanced against the distance to the stope in terms of gravity feeding from the dam, in terms of available head versus pressure losses for the required flow rate, and in terms of the pressure rating of suitable piping. Feeding the stopes from the dam can involve as many dedicated lines as required. Each of these ranges can feed to a different level or area as long as the dam height can drive the backfill to the paddock.
The advantages of the pressure break or underground dam system are:
- There is less space constraint in the main shaft.
- Pipeline pressures in the shaft are kept to a minimum.
- With only one pipeline in the shaft any failure can be more easily and rapidly dealt with resulting in less shaft time being required.
- Piping requirements are reduced.
- Blockage is less likely.
The disadvantages are:
- Further operator control underground,
- Actual control of the dam and agitation of the solids,
- Inability to feed stopes unless there is sufficient gravity head,
- Difficulty in controlling flow rates due to the short horizontal distances to the dam, and
- If the pipe in the main shaft fails larger volumes of backfill will fill the shaft due to the higher volumes of backfill transported down a single larger pipe.
Van Der Walt (1988) lists 10 points to consider when designing a backfill system. These points, although written in the context of free fall systems, also apply when considering the design of a full flow system:
- The transport velocity of the backfill must be significantly higher than the deposition velocity to prevent the backfill from settling out.
- The transport velocity must be kept as low as possible to minimize friction losses and pipe wear.
- Standard pipe sizes are preferred.
- In vertical columns, the maximum flow rate is at the point where the frictional losses exactly equal the available potential head.
- The flow rate of backfill through the system is determined by the inlet conditions only, i.e. the manner in which, and quantity of, backfill delivered to the vertical column.
- Maximum working pressures in the system will be found at the bottom of vertical columns, and determined by the frictional losses in the horizontal columns because this is an open-ended pipe system with free fall conditions in the shaft columns.
- Experience at pilot plants in the Goldfields of South Africa Group has shown that for practical implementation of the system each contractor should be allocated his own backfill range, if possible and where contractors are used.
- Usually backfill shaft columns have to be installed in shafts that are already equipped. However, this limits the available shaft space.
- Bursting discs should be provided at the points of maximum pressure in case of blockage in the pipe.
- Provision must be made for the flushing of lines before and after filling.
Step by step design procedure
A step by step procedure for backfill system design is given here to ensure that data from each step is available for the subsequent step.
Step 1: Determine the backfill flow rate
This is provided for in two ways, either as a predetermined flow rate or as a calculation based on the stope size and the mining schedule, to ensure that the cavity is filled and stable before any adjacent stopes are opened. Flow rates vary from 25m3/hr for narrow tabular operations to 200m3/hr for massive operations.
Step 2: Determine the static head
The static head is the pressure available owing to gravity under vertical conditions. It is calculated by taking the change in elevation of the backfill column between the shaft collar, or dam level, and the level that the column feeds. It is calculated as:
ΔH = pgh
where ΔH is the static head in kPa
p is the relative density of the mixture
g is the gravitational constant, 9.81 m's2
h is the difference in vertical elevation in m
Step 3: Determine the total pipeline length
This is simply the total length of the pipeline from the shaft collar, or underground dam to the area to be filled. It includes both vertical and horizontal piping.
Step 4: Piping selection
Pipe selection is based on the pressure rating of the available piping to an acceptable code of practice, such as ASME B31.3, and on the velocity through the pipeline to achieve the required flow rate. The backfill transport velocity is calculated by dividing the required flow rate (m3/s) by the internal cross-sectional area of the pipe (m2). Ideally, pipes are selected with backfill velocities between 2 and 6m/s for settling backfills and 1.5 to 3m/s for non settling backfills. Essentially, the objective is to ensure that no settling of large material occurs. Thus, pipe selection is an iterative process in relation with the pipeline frictional losses.
Step 5: Determination of frictional losses
Once the initial proposed pipe size/s have been selected, the pressure gradient or friction losses, for a range of velocities/flow rates, at the required backfill relative density, need to be determined for those pipe sizes in order to balance the system. This data has to be generated in a closed loop pipeline system such as shown in Figure 3:
A full flow system is described by Thomas et al (1979) as one in which “the system operates at the velocity at which the system pressure losses, due to pipe friction, bends, entry and exit orifices, are just balanced by the gravity head due to the difference in elevation between the discharge point and the free surface of the fill in the agitator tank". This balanced system full flow system, is one where the total pipeline frictional losses, calculated by multiplying the pipe line pressure gradient, at the specific conditions of backfill velocity and backfill relative density, by the total pipeline length, must equal the available static pressure.
Should the total frictional losses not equal the available pressure head a number of options are available to change the frictional loss balance:
- Change the pipe diameter. Decreasing the inside diameter will increase the frictional losses and increasing the pipe inside diameter will decrease the frictional losses.
- Increase the slurry relative density. This will increase the frictional losses; however, this also increases the static pressure requiring increased pipe pressure ratings.
- Utilize some form of energy dissipation.
In all cases the action taken on pipe will have ramifications on wear, flushing and pipe specification. Balancing the backfill system involves determining whether more friction or less friction is required in the system and then undertaking an iterative process of evaluating different pipe diameters to achieve a combined total pipeline friction loss that equals the available static pressure. This will typically result in different pipeline velocities through the different pipe sizes required to balance the system.
In selecting different pipe sizes to balance the system, the designer must be cognizant of a number of issues:
- Backfill velocities, both too high and too low a velocity can lead to either high wear rates or settling.
- Operating pressures experienced in increasing the system frictional losses, generated by a smaller bore pipeline, may lead to upstream pressures on the piping that exceed the initial pipe size pressure rating.
- Standard pipe sizes are always preferable.
With the pressure break system the requirements are to:
- Transport the total backfill requirement to an underground dam prior to backfilling of the actual worked stopes.
- Transport the backfill to an underground dam at a rate greater than the rate at which the dam is emptied.
This ability to fill the dam rapidly and yet maintain acceptable backfill velocities in the pipeline means that large bore piping is required. The decision on the size of pipe to use will be determined by:
- The number of ranges supplied by the dam and their flow rates. If the dam has 20 ranges feeding from the dam at 25 m3/hr the supply pipeline from the surface should ideally supply backfill at a flow rate not less than 500 m3/hr.
- The backfill velocity within the supply pipeline should return a wear rate that is acceptable. Large bore high pressure pipelines are expensive and ideally should last the lifetime of the mine.
- The ease with which the piping can be obtained along with fittings, etc.
- The amount of energy to be dissipated to ensure full flow conditions within the pipeline. This point will have to be balanced in respect to the lifetime of a small bore pipe running open ended at high backfill velocities with no energy dissipation, and that of a large bore pipe running full flow at low velocities but requiring energy dissipation.
Two pipelines must exist in the pressure break system, i.e. two surface to dam pipelines must be available to ensure backup should a failure occur. This also means that, if necessary, the pipelines can be designed to deliver less solids and if greater quantities of backfill are required by the dam the second, or backup, line can be used.
The necessity for energy dissipation has been shown in the previous discussion. The basic principle of energy dissipaters (EDs) is to increase the total systems pressure losses substantially, by forcing the backfill to flow through some form of restriction, or follow a convoluted path. The result is a high pressure zone at the inlet side of the dissipater and a zone of lower pressure at the outlet side. The pressure dissipated by the ED will be acting on the pipe wall immediately upstream of the ED. The Energy dissipater can be used to limit backfill velocities in pipelines through increasing the pipeline frictional losses over a short distance. This results in decreased wear rates.
Start-up and shut-down for pastefill systems
Paste backfill operations have a few characteristics associated with the large bore piping that is generally used and, must be catered for.
On starting up a typical paste backfill system the pipeline is slicked. This involves flushing the entire pipeline from the surface with water or a mixture of water and cement. This procedure is used in the pipeline transport of concretes to prevent water being removed from the leading core of concrete to “wet” the pipe walls. Simply put, the dry pipe walls “absorb” a certain amount of water from the concrete that subsequently “sticks” to the pipe wall providing the boundary lubrication to the following concrete. If this occurs, the leading core of concrete dewaters and a plug of dry material forms. A direct particle-particle interaction will then take place between the aggregates in the concrete, leading to interlocking of the particles and a blockage.
In the case of typical small bore backfill pipelines where the pipeline diameter is seldom greater than 80NB, most commonly 50NB, it is a simple task to fill the pipeline with water and follow with the backfill. Backfill relative densities are typically between 1.70 and 1.80. This correlates to between 66 and 71 mass % solids. The flow rate of water in an open ended 80NB pipe with a vertical length of 1000 m is approximately 160m3/hr and that of the backfill at a relative density of 1.70, approximately 138m3/hr, depending on the situation. This water flow rate can be readily achieved as flushing is generally only carried out for 20 minutes or less. The relatively small diameter of the pipeline and the similarity in the flow rates results in the water and backfill remaining in contact the entire length of the pipeline. In fact, the water and backfill mix probably resulting in the gradual change in flow rates between the water and the backfill and no impact loading of the pipe walls from backfill falling under gravity in an empty pipe.
Paste backfill pipelines are 150NB or larger. A vertical distance of 1000m will require a flow rate of approximately 960 cubic metres of water every hour to keep the pipeline full. The water will be travelling at a velocity of approximately 16m/s in a 150NB pipe. With the same assumption of 1000m of vertical pipeline and a static head of 23MPa, for a typical paste backfill system, the paste backfill will need to dissipate 23kPa/m of pressure to return a full flow system. Comparing this to the water flow rate and velocity, it is obvious that the water leaves the pipe long before the paste backfill has travelled any distance. This will result in the paste backfill falling under gravity and impact loading of the pipe walls increasing wear rates. This situation will be further exacerbated if the paste backfill flow rate is not maintained at 151m3/hr as a free fall situation will then exist in the system. It is this “free fall” effect that results in rapid backfill pipeline failures and blockage. The pipe wear is seen as long striations in the steel and in worse cases, the pipe is completely removed and only tendrils of steel exist between the last section of pipe and the subsequent section of pipe. Figure 4 shows a typical example of free fall wear in a backfill pipeline.
Figure 4: Typical free fall pipline wear
Furthermore, cyclic loading of the pipe wall at the paste/vacuum interface, as falling slugs of paste impact the interface occurs can lead to one of two effects:
- Dewatering of the paste at this interface and possible blockages occurring thereafter.
- Dynamic loading on the pipe wall owing to the pressures exerted by the water layer on compression by the falling slugs of paste, possibly exceeding the pipe tensile strength or the fracture strength of country rock in unlined boreholes.
All these effects occur on the start up of a paste backfill system and can lead to pipe failures or blockages. In order to introduce paste backfill into the pipeline without suffering from free fall, or “air mailing,” it is necessary to have the pipeline full of another medium, ideally water, before introducing the paste backfill. Then the other medium must be drained off at a flow rate similar to that of the paste. In this manner the paste backfill can be brought to the bottom of the vertical section without any high velocity scouring of the pipe walls.
In order to achieve this, the pipeline must be sealed off in the horizontal such that no vertical sections exist downstream. The valve (unit) used to close the pipeline must, when opened, have a bore the same as that of the pipeline. The unit must also be able to carry the total static head of the full column of water. The unit needs only to operate in the fully closed or fully open situation. Associated with this unit are smaller high-pressure valves also capable of withstanding the full static head in the pipeline. It is envisaged that a pinch valve with a high pressure rubber lining would be used as the main pipeline shut-off, although owing to the operation of the valve, either open or closed, any high pressure valve system, with a smooth bore the same diameter, when opened, as the pipeline, could be used. The smaller valves are positioned upstream from the pipeline closure unit. It is through these smaller valves that the water is drained from the pipeline. Provision must be made to run the water off through a pipeline to a drain. When the paste enters the horizontal piping after the vertical section, the main shut pipeline off valve can be opened and the smaller valves closed. The positioning of this system in the pipeline must ensure that the remaining horizontal section to the stope is “slicked” with the remaining water. This can be achieved by putting the system at the end of the line or ensuring that when the system is opened sufficient water remains for slicking the remaining line. The main criteria on opening the system is that the paste must have reached the bottom of the vertical transport distance.
This constitutes the system for starting up the paste backfill system without damage to the pipeline or threat of blockage. This system does not, however, ensure whether the overall system is in full flow or free fall.
Bain, A.G. and Bonnington, S.T. (1970). The hydraulic transport of solids by pipeline. Pergamon Press.
Barker, M.L. and Truscott, B. (1974). The development and operation of a test facility for pipeline abrasive wear measurement. Colorado, U.S.A., Paper J3.
Carter, A.M.F. and Howarth, H. D. (date). A literature review of the factors that influence the wear of backfill handling systems.Report No. M319. Council for Mineral Technology.
Cooke, R. (1991). Performance of backfill distribution systems at high solids concentrations.Hydraulic Conveying Solutions, Report No. TDS91012, August. Gold and Uranium Division, Anglo American Corporation.
Cooke, R. (1991). The dense phase hydraulic transport of high concentration cyclone classified tailings in pipelines. Ph.D thesis. University of Cape Town.
Cooke, R. (1992). The influence of binder addition on the hydraulic transport properties of backfill slurries. Hydraulic Conveying Solutions, Report No. TDS92014, August. Gold and Uranium Division, Anglo American Corporation.
De Bree S.E.M., Rosenbrand, W.F. and De Gee A.W.J. (1982). On the erosion resistance in water-sand mixtures of steels for application in backfill pipelines. Hydrotransport 8. Stephens H.S. (ed), BHRA fluid Engineering, Cranfield.
De Figueirdo, F. (1987). Design of deslimed tailings backfill system. Internal report, Technical note No. T 27/87, Task No. GD3VO1. Chamber of Mines Research Organisation.
Epithite, H.J. (1985). Rubber lining - The soft option against abrasion. Bulk Solids Handling, 5:5 Oct. pp.1041-1047.
Gilchrist, I.C.R. (1988). Predicting the values of flow parameters for the design of pipelines conveying backfill slurries.Backfill in South African Mines. South African Institute of Mining and Metallurgy. pp.235-261.
Guest, P.J., Lander, W.K. and Rabenstine, R.B. (1986). Conceptual design of a deslimed tailings backfill distribution system. Research Report No. 28/87, Project Number No. GD3R09. Chamber of Mines Research Organisation.
Hinde, A.L. (1989). Implications of backfill compositional changes. The Alleviation of Classified Tailings Backfill System Problems, Backfill Seminar, Chamber of Mines Research Organisation, September.
Paterson, A.J.C. and Lazarus, J.H. (1989). Backfill pipeline tests for Cooke III shaft – Randfontein. Hydrotransport Research, Report No. JCI 89 161. Johannesburg Consolidated Investments.
Petersen, J. (1988). Wear in backfill backfill pipelines - A review of the present knowledge and a proposed strategy for future research. Technical Note No. T39/88, Task No. GD3R12. Gold Exploration Laboratory, Chamber of Mines Research Organisation.
Sauermann, H.B. (1982). The influence of particle diameter on pressure gradients of gold slimes pumping. Proceedings Hydrotransport 8. BHRA Fluid Engineering, Cranfield, U.K.
Schmidt, E. and Spearling, A.J.S. (1990). Internal Backfill Guideline Report. Technical and Development Services, Gold and Uranium Division, Anglo American Corporation.
Shook, C.A., Haas, D.B., Husband, W.H.W. and Small, M. (1981). Relative wear rate determinations for backfill pipelines. Journal of Pipelines, No. 4. pp. 273-280.
Spearing, A.J.S. and Steward, N.R. (1992). Backfill, not paste, will be the fill of the future. Canadian Mining Journal, June. pp. 31-40.
Steward, N.R. (1990). The determination of wear related parameters for nine classified tailings. Gold and Uranium Division Report: April. Anglo American Corporation of South Africa.
Steward, N.R. (1991). The determination of the relationship between pipe wear and pipe diameter. Gold and Uranium Division Report: February. Anglo American Corporation of South Africa.
Steward, N.R. (1991b). The determination of wear relationships for two classified tailings at high relative densities. Gold and Uranium Division Report: July. Anglo American Corporation of South Africa.
Steward, N.R. (1991c). The determination of pipeline wear relationships for Fosroc Filset Binder modified classified tailings at high relative densities. Gold and Uranium Division Report: December. Anglo American Corporation of South Africa.
Steward, N.R. (1992). The determination of pipeline particle degradation and the wear related parameters for Elandsrand cyclone classified tailings and President Steyn full plant tailings. Gold and Uranium Division Report: April. Anglo American Corporation.
Steward, N.R. and Heckroodt (1989). Polyurethane as a wear resistant material in backfill handling applications. Hydrotransport 11: October. Stratford-upon-Avon, UK. pp.523-530.
Thomas, E.G., Nantel, L.H., Notley, K.R. (1979). Fill Technology in Underground Metalliferous Mines. International Academic Services Limited. Ontario, Canada.
Truscott, G.F. (1975). A literature survey on wear in pipelines. BHRA Fluid Engineering publication, TN 1295.
Van der Walt, P.G. (1988). Reticulation systems for deslimed tailings. Backfill in South African Mines. South African Institute of Mining and Metallurgy. pp. 393-403.
Wang Q. (1986). Wear resistance of steels under wet-abrasive erosion conditions. Wear:112. pp. 207-216.
Warman Technical pamphlet, Mitchel Cotts Engineering (Pty) Ltd.
Wingrove, A.C. (1989). JCI experiences with vertical "free fall" and "full flow" backfill systems. Backfill Transport in the Mining Industry, Current Technology, September. Hydraulic Conveying Association of South Africa. S.A.