The use of cemented paste backfill in underground mining is commonplace as a method for tailings disposal, rockmass confinement or minimising convergence between the hanging wall and foot wall of the operation. The application of a cemented paste fill can also enable greater extraction ratios within certain deposits, as ore that would ordinarily be left as pillars can be mined and replaced by a paste fill.
Paste fill generally comprises a filler material (usually fine tailings), water and binder. In many operations the binder is Portland cement, and represents the largest single cost of the backfill.
When optimising backfilling costs, the first areas to be investigated are the binder contents and curing periods. In addition to modifying the binder and curing periods alone, there are opportunities to modify the backfill based upon its rheological properties and still achieve the design intent. This article discusses the rheological approach which can often be overlooked.
Strength and water to cement ratio
The final strength achieved of a paste backfill is significantly influenced by the water:cement (W:C) ratio of the paste recipe, with the resultant strength of the backfill decreasing as the water content increases. Therefore, to minimise the cost, a paste should contain the minimum quantity of water possible based on the constraints of reticulation (the rheological approach) to maximise the W:C ratio and allow lower binder doses. Typical measured data and results are shown below.
The Role of Rheology
In paste backfill systems, the paste is typically delivered to the stope via a borehole/pipe network referred to as the underground distribution system (UDS). An important consideration then is the rheology of the paste, as at higher mass (and volume) concentrations the yield stress and viscosity of the paste typically become increasingly sensitive to minor changes in water concentration, as shown in the figure below.
Consequently, the paste must contain sufficient water to flow to the chosen stope, while simultaneously minimising the water:cement ratio. The relationship is a balancing act; the minimum rheology necessary to allow the paste to flow to the stope that will achieve the maximum strength in the minimum time.
Targeting Paste Recipes
The spatial variation of stopes within the mine means that in each case the pipe run will lead to different static heads and differing friction losses to deliver paste to the stope. By developing a thorough understanding of the flow characteristics of the paste backfill the paste recipes for individual stopes throughout the mine can be modified on a stope-by-stope basis to minimise the cost of backfill. This analysis should be undertaken during detailed design of a mine backfill system.
If time does not permit the development of unique recipes for every stope, zoning can be applied to optimise the W:C ratio in the paste for different areas in the mine. The zoning is characterised by differences in pressure head and frictional losses, where the head increases with depth in the mine, and the total frictional losses typically increase in distance from the borehole. The following figure shows different zones that can be defined based on the available head and frictional losses anticipated and is useful both to illustrate the concept and also for high level assessments or studies.
This approach will lead to some of the stopes within each zone being overdosed with binder, but this still represents a marked improvement over using a general ‘one-size fits all’ approach for the whole mine, in which entire zones may be overdosed. Narrow conical angles require high friction losses that can be achieved by minimising the W:C ratio and wide conical angles result in higher W:C ratios to decrease friction losses.
Benefits of Binder Optimisation
The following table summarises the theoretical potential for savings associated with rheology and binder optimisation, assuming a target strength of 800 kPa at 28 days throughout the mine.
The non-optimised solution is a fairly typical approach of having one binder dose for all stopes, and thinning the paste recipe to reach the furthest areas in the mine. This leads to overdosing in the stopes closer to the paste plant.
The optimised solution uses an understanding of the rheology of the cemented backfill. Increasing the mass concentration (decreasing the water content) is planned strategically and based on what can flow to the selected stoping zone. The reduced water content allows a lower binder dose to be used to achieve the target strength.
- Binder consumption is typically the major operating cost for backfill production
- A thorough understanding of the rheology of the paste and the mine UDS allows for manipulation of the binder recipe, to target the thickest paste that will reliably flow (or be pumped) to the stope
- By reducing the water content in the paste recipe, the binder dose can also be reduced to maintain the same water to cement ratio leading to significant operational savings, both annually and over the life of the mine
- To quantify savings, a banding approach has been presented, splitting the mine into zones based on the difficulty of reticulation, (considering flow path, static head and distance)
- Whilst the design and use of bespoke paste recipes may increase the technical workload for the backfill engineer, the cost savings readily justify the additional effort. Paterson & Cooke develops and uses our own tools regularly to improve the modelling accuracy
About The Author
CEng, MEng (Chemical)
Damian is a chartered Chemical Engineer and a Member of the Institute of Chemical Engineers. He has led a wide variety of industrial Mining projects specialising in hydraulic modelling and plant design. Damian has a strong knowledge of process design and review including HAZOP chairing. He also has extensive project management experience having worked as Lead Engineer from pre-feasibility through to construction, commissioning and hand over on numerous Capex projects. Damian joined Paterson & Cooke in February 2015 and is based in the Cornwall office in the UK.