Hydraulic Conveying Association of South Africa 4th One Day Seminar on Hydraulic Transport in the Mining Industry 7 April 2003, Indaba Conference Centre

How to Reduce Pipe Friction in Slurry Flows Dr Nigel Heywood Aspen Technology, Harwell, UK

Abstract This paper describes three main methods for reducing the frictional pressure loss, and therefore the pump discharge pressure requirement, when transporting viscous, often nonNewtonian, slurries and pastes in pipelines. The first method reduces the degree of flocculation of the particles in the slurry using suitable chemical additives, and thereby reduces the slurry viscosity. The second method makes use of boundary liquid (such as water, oil, or polymer solution) which is injected at comparatively small flowrates into the pipe downstream from the pump to form a lubricating annulus adjacent to the pipe wall. The third method involves gas injection into the pipe downstream from the pump to form a slug flow pattern which results in substantial frictional pressure loss reductions for shear-thinning, nonNewtonian slurry, initially flowing in the laminar flow regime prior to gas injection. 1.

Introduction

Many new developments have taken place over the last few years in slurry handling practice (Heywood, 1999), and particularly in slurry pipeline technologies (Heywood & Alderman, 2003). One example of this is the increased tendency for slurries and pastes containing very fine particles to be pumped at increasingly high concentrations. These slurries tend to be pseudo-homogeneous and do not segregate readily under gravity forces. They are often highly viscous and exhibit highly shear-thinning, non-Newtonian flow properties. Figure 1 indicates the typical consistency of these slurry types.

Figure 1: Viscous, Fine Particle Slurry Discharging from a Pipe

In addition, it is also possible to pump relatively low moisture, unsaturated (compressible) filter and centrifuge cakes (Figure 2) using both reciprocating positive displacement pumps and some classes of rotary pump using bridge breakers and single or double intermeshing, contra-rotating augers at the base of feed hoppers.

Figure 2: Unsaturated Sewage Sludge from a Compression Cake Filter In order to pump both classes of slurry (either saturated paste or unsaturated cake) high pump discharge pressures are frequently required. These discharge pressures result in: •

High pumping power consumption

•

High pump differential pressures, leading to pump slippage and higher wear rates in some pump types

•

More expensive, thicker-walled pipe

•

Less flexibility to pump slurry at higher, more viscous solids concentrations

There are various techniques to reduce the frictional pressure loss in pipe flow, which is often the major contribution to the pump discharge pressure. This paper describes three of these methods: 1. Reduction in the level of particle flocculation through the use of additives to reduce the zeta potential on particle surfaces; 2. Injection of water or some other liquid (such as oil or aqueous polymer solution) into the pump discharge pipe using an appropriate injection ring, thus creating a lubricating annulus in the pipe for the viscous slurry; 3. Injection of air (or inert gas) into the pump discharge pipe to reduce the proportion of the inner pipe wall wetted by viscous slurry.

2.

Use of Additives to Disperse/Deflocculate Fine Slurry particles

Adding soluble ionic compounds to flocculated slurries (i.e. slurries where a significant proportion, say 5 to 10%, of solids are below approximately 5 µm) can result in substantial pressure drop reductions in pipe flow in the laminar flow regime. However, the effect is usually not discernible in the turbulent regime. The ionic compounds disperse the particles, thereby breaking up the flocculated particle network that gives rise to higher shear stresses in pipe flow. In order to disperse particles, the charge on their surfaces needs to be of the same sign, and the higher the charge the greater the repulsive forces between particles. It is only relatively recently that the reduction in pressure drop has been related to the degree of deflocculation in quantitative terms. Several studies (Horsley et al, 1980 & 1982a, b) have been undertaken in which the zeta potential on particle surfaces, defined as the potential in the electrochemical double layer at the interface between a particle moving in an electric field and the surrounding liquid, has been measured and correlated with the pressure drop reduction. The zeta potential is a measure of the repulsive forces acting on the particles and thus is useful in measuring the dispersing effects of chemical additives. Figure 3 shows some typical head loss-slurry velocity data (Horsley & Reizes, 1980) for the flow of a 43% by volume sand slurry (37% by weight of particles less than 10 µm) at different pH levels in the range 6.2 to 10.8; the pH was controlled by the addition of nitric acid and sodium hydroxide. The changes in pressure drop in the laminar regime were observed to be greater as the sand concentration was raised.

Figure 3: Effect of pH and Zeta Potential on 43% Volume Concentration Sand Slurry (d50 =17 µm) in Pipe Flow (Horsley & Reizes, 1980)

Figure 4 shows a plot of head loss against zeta potential. This illustrates that the higher the zeta potential (i.e. the larger the negative value) the lower the pressure drop.

Figure 4: Effect of Zeta Potential on Head Loss Gradient for 33% and 43% Volume Concentrations of Sand Slurry (Horsley & Reizes, 1980) The effects of chemical additives have been studied (Sikorski et al, 1982)) on a number of mineral slurries including drilling muds (thinners such as sodium acid pyrophosphate and sodium hexametaphosphate), phosphate rock slurries (caustic soda), limestone cement feed (combination of sodium tripolyphosphate and sodium carbonate), and coal slurry (sodium tripolyphosphate, sodium dioctyl sulphosuccinate and sodium carbonate). Various alkaline additives have been used to modify the flow properties of coal-water slurries (Shook & Nurkoski, 1977 & Teckchandani & Shook, 1982). Several thinners have been investigated (McInnes, 2002) at different concentrations to reduce the viscosity of chalk slurry, so facilitating long distance pumping, and additives have been used to reduce pumping energy requirements for flyash waste slurries from power stations (Thomas & Sobota, 2002). The level of dispersion of a flocculated slurry, and therefore indirectly its viscosity level, can now be assessed on-line through zeta potential using commercially-available instrumentation. Off-line zeta potential measurements have been possible for some time. With the on-line instrument, an electroacoustic sensor (Hunter, 1998 & 2001) applies high-frequency alternating voltage pulses across the slurry. This causes the slurry particles to oscillate at a velocity that depends on their size and electric charge (or zeta potential). This particle motion generates sound waves, which in turn depend on particle size and zeta potential. The input voltage pulse is applied across two flat, parallel platinum electrodes that are in contact with the slurry. The slurry flows vertically upwards through a polyphenylenesulphide (PPS) spacer in the sensor. One of the platinum electrodes is coated onto a rectangular glass delay line, at the opposite end of which is mounted a thin ultrasonic transducer. The gated amplifier produces a pulse of high frequency alternating voltage of about 3 µs duration, which is applied across the parallel electrodes. The resulting high frequency sound wave produced by the motion of the slurry particles in the alternating electric field travels down the glass delay line and is detected by the ultrasonic transducer.

The measured electrical signal, called the ESA, contains information about the particle size and charge, and is the primary measurement made by the electroacoustic sensor. The ESA measurement is repeated a number of times at a range of frequencies from 1 MHz to 20 MHz. The Fourier Transform of these signals is referred to as the ESA frequency spectrum. From this spectrum, the particle size distribution and the zeta potential are ultimately calculated. 3.

Injecting Boundary Layer Liquid into Pipe

Very highly concentrated slurries can now be pumped using either progressive cavity or reciprocating piston pumps (Zey, 1999) by force-feeding the slurry into the pump body using various auger designs and bridge breakers at the base of wedge-shaped hoppers. Even unsaturated solid/liquid/air mixtures such as compressed filter or centrifuge cakes can be pumped in this way. To reduce the frictional pressure loss in the pump discharge pipe, it is now common for many pump manufacturers to offer boundary layer injection facilities whereby a liquid is injected at three or four points through the pipe wall to generate an annulus that helps to lubricate the flow. The liquid injected is often water but greater reductions in frictional pressure loss can be achieved using aqueous polymer solution, waste or heating oil or polyelectrolytes. Because the central slurry core is so concentrated and viscous, mixing between the annular wall layer of injection liquid and the slurry can be minimal so the friction reduction effect is maintained over a significant pipe length. Figure 5 shows a liquid injection arrangement supplied by Schwing America for their reciprocating piston pumps, and Figure 6 shows how a lubricating annulus is formed in the pipe.

Figure 5: The Boundary Layer Injection Method used by Schwing Pumps

Figure 6 : Details of the Schwing Pumps Boundary Layer Injection Method

Figure 7 shows the extent of the reduction in pump discharge pressure for a progressive cavity pump which is pumping sewage sludge at various solids concentrations, and using water at two different concentrations of polymer solution. It can be seen that a 0.5% polymer solution injected at 20 litre/h into the pipe carrying sludge at 2 m3/h (1:100 flowrate ratio) gives rise to the greatest reduction in discharge pressure from 13 bar down to about 1.8 bar. Other conditions also give substantial reductions in discharge pressure.

Figure 7: Reductions in Pump Discharge Pressure for Sewage Sludge using Alternative Boundary Layer Liquids (Courtesy, Seepex Pumps)

The nature of the boundary layer fluid appears to have a significant effect on the pressure drop reduction on the pipe wall. Various liquids are used by Putzmeister Pumps depending on the application (Zey, 1999). Table 1 from Putzmeister shows the various liquids they use.

Table 1 : Boundary Layer Fluids used by Putzmeister Pumps Boundary layer fluid Water

Advantages Cheap

Heating oil/waste oil

Greater cost benefit when used as combination aid High efficiency

Polyelectrolytes PMLC

Extremely efficient Does not mix with slurry

Disadvantages Mixes with slurry/sludge Used only in incineration plant

Pressure Reduction 20-50%

Mixing station may be required Mixing station required

50-75%

25-50%

70-90%

Figure 8 shows the substantial effect that either water injection or other special lubricant can have on the pressure gradient (Zey, 1999).

Figure 8: Putzmeister Pumps Data for Pressure Gradient Reduction for the Flow of 37 to 47% Sewage Sludge

4.

Injecting Air Downstream From Pump

It has been known for several decades that injecting air (or another gas) into a shear-thinning slurry in laminar pipe flow will result in a reduction in frictional pressure loss. No such effect occurs for the laminar flow of a Newtonian slurry, or for the turbulent flow of a slurry having any specific laminar flow property. Figure 9 shows some typical data; the drag ratio in this figure is defined as the ratio of frictional pressure gradient along the pipe with gas injection to that with no gas injection, at constant superficial slurry velocity, i.e., constant slurry flowrate.

Figure 9: Typical Reductions in Frictional Pressure Loss for the Horizontal Pipeflow of Shear-thinning Slurry at Various Superfical Slurry Velocities (Farooqi et al, 1980)

The friction reduction effect occurs in both horizontal (Heywood & Richardson, 1978, and Farooqi et al, 1980) and vertical (Heywood & Charles, 1980, and Khatib & Richardson, 1984) slurry pipe flows, and there is the additional advantage of reduced static head in vertical pipe flows. Figure 10 shows how the gas can distribute itself within the slurry in both pipe orientations, giving rise to six identifiable flow patterns in horizontal pipe flow, and four in vertical pipe flow. In horizontal pipeflow, frictional pressure loss reduction occurs for practical purposes mainly in bubble and plug or slug flow (intermittent flow), but it can occur in stratified flow at low slurry flowrates (Heywood & Charles, 1979). For a given flowrate combination of gas and slurry in a pipe, it is possible to predict the resultant flow pattern. This information is often presented in the form of a flow pattern map. An example of one developed for gas and non-Newtonian slurry flow in a horizontal pipe is given in Figure 11.

Figure 10: Flow Patterns for Horizontal and Vertical Gas/Slurry Pipe Flow

Figure 11: Flow Pattern Map for Gas/Slurry Flows in Horizontal Pipes (Chhabra & Richardson, 1984) Friction reduction is greater the more shear-thinning the slurry and a maximum effect occurs when the combined gas and slurry pipe flows nears the laminar-turbulent transition for the slurry flow. In this way, the maximum effect can be reliably predicted from a knowledge of the slurry rheological properties, the pipe diameter and the gas and slurry flowates. Figure 12 shows the maximum reduction in frictional pressure drop as a function of the pipe Reynolds number for slurry flow alone.

Figure 12: Maximum Reduction in Pipe Friction (Minimum Drag Ratio) for a Power Law, Shear-thinning Slurry with various n’ and K’ Parameter Combinations (Farooqi & Richardson, 1982) The maximum reduction in pipe friction can also be correlated using a dimensionless factor, J, which is defined as the ratio of the superficial slurry velocity in the pipe to the critical value for laminar flow breakdown, raised to the power of (1-n). The parameter, n is the exponent in the power law model. This correlation is shown in Figure 13 which is particularly useful because it allows estimation, a priori, of the minimum achievable drag ratio from a knowledge only of the slurry properties (density and power law parameters) and the operating conditions (pipe diameter and superficial slurry velocity, Vsl. The corresponding gas velocity id calculated as the difference between the critical slurry velocity for laminar flow breakdown, (Vsl)c. and Vsl.

Figure 13: Correlation of Maximum Reduction in Pipe Friction using Gas Injection (Farooqi & Richardson, 1982)

There are many advantages to using air injection: (1) reduced discharge pressure requirement for a slurry pump for a given slurry flowrate though a given discharge pipe length; (2) increased capacity of an existing pipeline carrying a given slurry while retaining the same pump system; (3) extension of an existing pipe run while maintaining the same discharge pressure; (4) application of an existing pump and pipeline combination to more viscous, shear-thinning slurry, while maintaining the same discharge pressure. (5) reduced pump differential pressure, and therefore reduced slippage in some pump types, with a corresponding reduction in pump wear. Surprisingly, the advantages of using air injection have still largely been overlooked by industry, although a Polish sugar factory takes full advantage of the technology when pumping waste molasses (Dziubinski & Fidos, 1992). Numerous examples of highly viscous, shear-thinning slurries being pumped through pipework occur in industry, e.g., red mud waste from alumina production, pulverised fuel ash slurry from power stations, titanium dioxide slurry, chalk and clay slurries, etc. Many of these operations could benefit from the appropriate application of air injection. 5.

Conclusions

This paper has described three alternative methods to reduce pipe friction for highly viscous, often non-Newtonian slurries. In addition to frequent economic advantages to be had by employing one of these methods, other benefits also accrue. These include lower wear rates and therefore maintenance costs for pumps, and reduced instances of pipe blockages. These methods also increase the flexibility of existing pipeline systems when there is a requirement to increase slurry concentrations, and therefore slurry viscosities. It is recommended that each of these three methods is considered whenever an existing pipeline needs upgrading, or a new pipeline system is to be designed. References Chhabra, R.P. & Richardson, J.F. (1984) “Prediction of flow pattern for the co-current flow of gas and non-Newtonian liquid in horizontal Pipes”, Can J Chem Engng, Vol 62, Page 449-454. Dziubinski, M. & Fidos, H. (1992) “An industrial installation for two-phase transportation of carbonation mud”, Zuckerindustrie, Vol. 117, No.8, Pages 631-633. Farooqi, S.I. & Richardson, J.F. (1982) “Horizontal flow of air and liquid (Newtonian and NonNewtonian) in a smooth pipe. 2. Average pressure drop Trans I Chem E, Vol 60, Pages 323-333. Farooqi, S.I., Heywood, N.I & Richardson, J.F. (1980) “Drag reduction by air injection for highly shear-thinning suspensions in horizontal pipeflow”, Trans I Chem E, Vol 58, Pages 16-27. Heywood, N.I. (1999) “Stop your slurries from stirring up trouble”. Chemical Engineering Progress, September, Pages 21-41. Heywood N. I. & Richardson, J.F. (1978) “Head loss reduction by gas injection for highly shearthinning suspensions in horizontal pipe flow”. In : Proc Hydrotransport 5, Paper C1. Organised by BHRA Fluid Engineering, UK.

Heywood, N. I. & Charles, M.E. (1979) “The stratified flow of gas and non-Newtonian liquid in horizontal pipes”. Int J Multiphase Flow, Vol 5, Pages 341-352. Heywood, N. I. & Charles, M E (1980) “Effects of gas injection on the vertical pipeflow of fine slurry”. In: Proc of Hydrotransport 7, Paper F5. Organised by BHRA Fluid Engineering, UK. Heywood, N. I. & Alderman, N.J. (2003) “Developments in slurry pipeline technologies”. Chemical Engineering Progress, April, Pages 100-107. Horsley, R.R. (1982a) “Viscometer and pipe loop tests on gold slime slurry at very high concentrations by weight, with and without additives”. In : Proc. of Hydrotransport 8 . Paper H1, Organised by BHRA Fluid Engineering, UK. Horsley, R. R. (1982b) “The relationship between zeta potential and head loss gradients for slurry pipe flow with varying pipe diameter and volumetric concentrations” J Pipelines, Vol 3, Pages 87-96. Horsley, R.R. & J. A. Reizes (1980) “The effects of zeta potential on the head loss gradient for slurry pipelines with varying slurry concentrations”. In : Proc. Hydrotransport 7. Paper D3, Organised by BHRA Fluid Engineering. Hunter, R. J. (1998) “Recent developments in the electroacoustic characterisation of colloidal suspensions and emulsions”, Colloids & Surfaces, Vol A141, Pages 37-65. Hunter, R. J. (2001) “Measuring zeta potential in concentrated industrial slurries”. Colloids & Surfaces, Vol A195, Pages 205-214. Khatib, Z. & Richardson, J. F. (1984) “Vertical co-current flow of air and shear-thinning suspensions of kaolin”. Chem. Eng. Res. Des., Vol 62, Pages 139-154. McInnes, M.-A. (2002) “Investigation into the effects of slurry thinners on the rheology of chalk slurry”. In : Proc of Hydrotransport 15, Banff, Canada, Pages 375-384. Organised by BHR Group Ltd, Cranfield, UK. Shook, C. A. & Nurkoski, J. (1977) “Effect of some alkaline additives on the viscosity of coal water slurries”. Can. J. Chem.Eng., Vol 55, Pages 510-515. Sikorski, C. F., et al. (1982) “The effects of viscosity reducing chemical additives on slurry rheology and pipeline transport performance for various mineral slurries.” In : Proc 7th Slurry Transport Association Conference, pp 163-173. Teckchandani, N. & Shook, C.A. (1982) “Effect of a consistency-reducing additive on coal slurry pipeline pressure drop”. J Pipelines, Vol 2, Pages 43-47. Thomas, A. D. & Sobota, J. (2002) “Influence of additives on energy loss in pipeline flow of flyash mixtures” In : Proc of Hydrotransport 15, Banff, Canada, Pages 329-343. Organised by BHR Group Ltd, Cranfield, UK. Zey, W. (1999) “Putzmeister high-density solids pumps: design and application”. 54 pages. Published by Putzmeister AG, Germany.