Hydraulic performance and erosion wear of centrifugal pumps

Abstract: Centrifugal pumps are widely used for medium- and short-distance solid hydraulic transportation through pipelines that require medium head and pump outlet. The performance characteristics and erosion wear of Centrifugal pump components are the most critical design and selection parameters. The improved performance reduces energy consumption, while the reduction in erosive wear increases service life. There is an ongoing effort to estimate the performance degradation of the pump in handling water with different types of solid particles and to find ways to mitigate it. Different relationships are proposed to evaluate the performance of the pump in handling slurries. Regarding corrosion of Centrifugal pump components, different techniques are used to identify the areas of greatest localized wear and propose ways to reduce it. In this paper, experimental and numerical studies conducted in this field are discussed. For an optimized pump design, the choice of pump material, the nature of the solid particles, the flow characteristics of the slurry and the concentration of the solids play an important role. This review summarizes recent advances in the evaluation of pump performance and pump wear characteristics. Centrifugal Pump design needs to be optimized to handle slurries based on performance degradation and expected service life. foreword The hydraulic transport of solids through pipelines in the form of slurry has been widely used as an effective and cost-effective method for transporting bulk solids. The pumping unit is one of the most important components in the system. Centrifugal Pumps are roughly divided into two categories, namely positive displacement pumps and rotary power pumps. Rotary power pumps are usually used for continuous conveying and have wider flow paths for smooth flow of solid particles. Centrifugal pump is a kind of rotary power pump, which is mostly used for medium and small distance transportation of slurry in many industries. Centrifugal pumps are usually improved and developed on the basis of conventional pumps. These improvements include enlarged flow passages to accommodate larger particles, use of robust impellers with fewer blades, special seal arrangements and appropriate wear-resistant materials to ensure longer service life [1]. The performance and wear characteristics of these pumps are often studied using laboratory test rigs. The actual field performance of these pumps is then predicted based on these experiments. Most of the experimental work reported in the literature to evaluate these properties has a test-bed arrangement similar to that shown in Figure 1. picture Figure 1: Schematic diagram of the experimental setup used for pump performance and wear studies [1] This paper discusses some important experimental and numerical work on centrifugal slurry pump performance and wear estimation to help colleagues understand the current state of slurry pump development. Experimental Research on Performance Evaluation of Centrifugal Slurry Pump The performance of centrifugal slurry pumps plays an important role in the trouble-free and economical conveying of solid-liquid mixtures in hydraulic conveying systems. As shown in Figure 2, the performance of these pumps with water decreases with the presence of suspended solids. To determine the performance of these pumps with slurries, either experiments with actual slurries or predictions using available relationships [2]. Relationships are typically established to estimate additional parameters defined as follows. picture Figure 2: Schematic diagram of the effect of slurry on the performance characteristics of centrifugal slurry pump [2] picture picture picture The head reduction factor (KH) is defined as HR, and the efficiency reduction factor (Kη) is defined as ER. For similar pumps, point-to-point data from a pump of one size can be transferred to a pump of another size or the rotational speed change of a centrifugal pump using the similarity law expressed as the following dimensionless parameter picture picture picture The applicability of these laws to centrifugal slurry pumps is also one of the researchers' main concerns due to the effects of suspended solids and the geometry of these pumps. The analysis of the effect of solid particles on pump performance was initiated by Fairbank [3]. He studied the performance of a 3-inch centrifugal pump in handling slurries and suspensions containing two types of sand, 34 μm and 800 μm in diameter. He observed that for very fine particles, the head developed for slurry is similar to that developed for water, but for larger particles the head is lower. He found that the drop in the head-flow characteristic curve at constant velocity increases with increasing solid particle concentration and particle size, while the flow rate at maximum efficiency is independent of concentration and particle size. The input power increases linearly with the increase in the specific gravity of the slurry. He developed an expression for calculating theoretical head using Euler's equation by assuming that the flow velocity of solid particles at the impeller outlet is higher than the velocity of water. He also reports that the law of similarity applies to slurry pumps with a narrow range of rotational speed variations. Experimental work by different researchers, namely Mano [4] on a 2-inch slurry pump, Hasegava et al. [5] and Sasaki et al. [6] on a 3-inch slurry pump, and Herbitch and Valentine [7] on a 4" slurry pump. Stepanoff [8] summarizes all slurry pumps that handle different solids and concludes that the performance of the pump depends on the rheological properties of the slurry, particle size and solids concentration. As the specific gravity of the slurry increases, the input power increases, and the similar laws of traditional pumps also apply to centrifugal slurry pumps. Wideroth [9] used cloud technology to calculate the flow rates of solid particles and water in a standard non-clogging centrifugal slurry pump and a model dredging pump, respectively. He injected a cloud of solid matter into the slurry and recorded the time it took to travel a certain distance. He observed that there was no slippage at the outlet of the pump, while at the inlet, solid particles slipped in. To test the performance of the slurry pump he used two pumps and thirteen sand and gravel samples. He observed a larger variation in the characteristics of low-speed pumps compared to medium-speed pumps, and attributed this phenomenon to differences in the frictional paths of the flow within the pump. He also conducted experiments to investigate factors affecting pump performance and proposed a relationship for estimating head loss using dimensionless parameters. Hunt and Faddick [10] used three different sizes of centrifugal pumps to process aqueous solutions of flake particles, examined the dependence of pump performance on its size, and reported that pump size affects head and efficiency. Vocaldo et al. [11] investigated operating point selection based on solid particle concentration. They report that the total losses in the pump are caused by the pressure gradients required by the carrier fluid, particle interactions, and keeping the particles in suspension. Using dimensional analysis, they developed a semi-empirical relationship to predict the head reduction factor, and determined the constants of the empirical relationship using experimental data on metal and rubber-lined pumps. They experimented with different sizes of sand and clay at two speeds, namely 1180 rpm and 1780 rpm. They observed that the head loss of the rubber-lined pump was higher than that of the metal pump due to the better energy absorption capacity of the former. They also observed that similar laws apply to slurry pumps. Burgess and Reizes [12] conducted experiments on pump performance at two speeds, 800 rpm and 1300 rpm, and reported the applicability of similarity laws to pumps. They developed a relation to predict the head reduction factor and reported the accuracy of their experimental results as ±5 %. Cave [13] also used experimental data to develop a relation for evaluating the head reduction factor. Widenroth [14] suggested to characterize solid particles with average particle size to estimate head loss. He conducted tests in three test loops of different pipe diameters at different speeds to evaluate pump performance for two impeller geometries and fifteen materials, mainly sand and gravel of different sizes. He reported the dependence of dimensionless head flow characteristics on slurry concentration and material type. Sellgren [15] evaluated the performance of a four-blade rubber-lined pump delivering ore and minerals at 760 rpm and 1140 rpm. Based on experiments, he reported that slurries with wider particle size distributions had less effect on internal flow characteristics than slurries with equal particle size, and that the reduction in head was independent of pump speed. He also observed that up to 20 % solids concentration (by weight), the reduction in head equals the reduction in efficiency, but at higher concentrations the reduction in efficiency is greater than the reduction in head. He proposed a relational formula for predicting the reduction coefficient of head, and the error of the experimental data is ±15%. Want [16] discusses that losses are not always the same during the life of the pump. He reported that in the first half of the pump's life, the reduction in head loss was due to the frictional effects of the slurry, while in the second half, the head loss was due to inaccurate casting, joint discontinuities and an increase in internal eddy currents. Neerken [17] evaluated the performance of settled and non-settled slurries. He gives the variation of the reduction factor with the average particle size for solid specific gravity between 1.1 and 6.0 and particle size between 0.01 and 0.8 mm. To develop a pump sizing method, Walker and Goulas [18] developed a computer program that generated information on losses, pump performance, and operating points based on solids characteristics, slurry rheology, pump size, and pumping system details. information. They also include a program to estimate the required Centrifugal pump input power. Johnson [19] studied the problems associated with centrifugal pumps handling non-Newtonian slurries. He observed a reduction in the pump Reynolds number for water used to handle solids, and reported that a pump Reynolds number < 105 was not suitable for pumping non-Newtonian slurries. Remisz [20] experimentally evaluated the performance of centrifugal pumps for handling small particle slurries (dmax < 0.2) at 1480 rpm and 1720 rpm. He found that as the density of the mixture increased, the head increased and the efficiency and flow at the optimum efficiency point decreased. He also developed a dimensionless flow and mass density relationship for mixtures to evaluate pump performance for solid-liquid mixtures. Walker and Goulas [21] evaluated the performance characteristics of centrifugal pumps handling non-Newtonian slurries of coal/water and kaolin/water mixtures. They observed instabilities in the pump system curve at low flow rates, as shown in Figure 3. Pump performance is affected by the apparent viscosity of the fluid, with the system curve intersecting the pump characteristics at two or three points at low flow rates, while at high flow rates the performance is reduced due to the plastic viscosity of the fluid. They also report that the law of similarity does not apply to pumps handling non-Newtonian slurries