�鶹��Ƶ

The Jhimruk hydropower project has a desilting basin designed to trap 90 per cent of particles greater than 0.2 mm. However, from the beginning the plant has faced problems of abrasion in the hydro-mechanical equipment, due to the unexpectedly high content of silt particles finer than 0.2 mm. Settling basins, improved plant operation, and hard coating has reduced wear on the turbines. As a result, the loss of energy production has been minimised.

The Jhimruk hydropower plant is a 12 MW run-of-river plant built and commissioned in 1994. Owned and operated by Butwal Power Company (BPC), Jhimruk is located in the Pyuthan district inthe Mid-Western Regionof Nepal. The project benefits from a 205 m net head caused by the water diversion from the Jhimruk river to the Madi River. Both rivers meet around 30 km downstream.

The project includes a weir in the Jhimruk River that diverts water to two parallel settling basins, before it is conveyed at a design discharge of 7.05 m3/s to the semi-underground powerhouse through a 1 km long headrace tunnel and a 250 m long penstock. The water is finally discharged in the Madi River through a short tailrace channel. See the project layout in figure 1.

The diversion weir has a curvilinear shape with 205 m overflow length. The crest elevation is at 738 masl. Following the conventional design criteria of similar projects in Nepal the settling basins, of dimensions 42 m long, 5.5 m wide and 7 m deep, were designed to trap 90 per cent of particles larger than 0.2 mm. The powerhouse hosts three Francis turbines of 4 MW each.

Hydrology and sediment

The catchment tributary to the Jhimruk weir comprises 645 km2, characterised by four months of rainy season (monsoon), lasting from June to September. The peak flow period occurs during July and August. The annual average precipitation is 1,610 mm, more than 80 per cent of which occurs during the monsoon period, and the mean annual inflow is 851 Mm3. The most significant sediment transport of around 80 per cent also occurs during the rainy season. In other Himalayan river basins the annual sediment load can reach 100,000 tons.

Sediment problems

Sediment data for the Jhimruk River was not available during the planning and design phase of the project. Two desilting basins were therefore designed based on general sediment references for Himalayan rivers and on conventional design criteria for hydropower plants in Nepal. The desilting basins were meant to trap 90 per cent of sediment particles larger than 0.2 mm in grain size.

The first sediment handling issues occurred after the first five months of operation, following the monsoon. Damage to the hydromechanical components (turbine blades, guide vanes, facing plates and casing) was so severe that it soon became evident that exposure to sediment abrasion through sand-laden water was much higher than originally anticipated.

Sediment abrasion by hard minerals is one of the most challenging aspects faced by hydropower projects in Nepal. It can lead to a reduction in the efficiency and design life of the turbines. The erosion suffered by the turbine wear and the guide vanes shown in figure 2 illustrates the high costs of operation and maintenance at Jhimruk.

Since commissioning in 1994, monitoring of suspended sediment at the plant has been carried out with a view to improving sediment management strategies. From 1994 to 1997, the monitoring programme recorded the suspended sediment concentration during the monsoon in Jhimruk River. The data shown in figure 3 presents mean values ranging from about 2,000 to 6,000 ppm, with maximum values ranging from about 20,000 ppm up to 60,000 ppm during peak flow. The runners were so damaged after each monsoon that the turbines needed to be repaired on an annual basis.

Particle size distribution and mineralogical studies were performed to complete the suspended sediment profile. Both are critical parameters for defining the potential abrasion in the turbines and for identifying the best sediment management strategies and the most appropriate hydromechanical equipment.

In 1996, a sample from the desilting basins showed that 90 per cent of the particles entering the turbines was finer than 0.1 mm in grain size, and that the mean diameter D50 was 0.025 mm. A more recent study in 2012 compared samples from both locations at the headworks above the weir and at the tailrace of the power plant. As shown in figure 4, the study found that 80 per cent of the sediment at the headworks had a grain size range between 0.1 and 0.2 mm, and about 95 per cent of the sediment at the outlet was within this size range. Both studies demonstrated that the settling basins filter sediment particles larger than 0.2 mm and that the critical size of sediment particles reaching the turbines is of 0.1 to 0.2 mm. Civil structures such as a desilting basin to trap sediment particles finer than 0.2 mm are extremely costly and therefore the exclusion of particles smaller than 0.1 mm is economically unfeasible.

The mineralogical studies found that about 80 per cent of the sediment is composed of two hard minerals: quartz and feldspar, whose hardness at Mohs scale is 7 and 6 respectively. These minerals will wear the uncoated turbines and their components away, as their hardness does not normally exceed a 5 in the Mohs scale.

The high percentage of hard minerals, combined with the large proportion of sediment particles finer than 0.2 mm that are not being trapped in the settling basins, are causing abrasion of the turbines. As a result, Jhimruk power plant has normally been operated at a lower capacity than is installed, including temporal plant shutdown.

Sediment management strategies

Two settling basins were designed to trap sediment particles greater than 0.2 mm in grain size. Despite the proven efficiency to exclude these particles, the settling basins did not prevent severe damage to the turbines. Figure 5 shows the desilting basins at Jhimruk headworks.

In 2003, Hydro Lab in Nepal launched a research project on optimum sediment handling strategies in run-of-river projects, featuring Jhimruk as its initial case. Measures have been researched to reduce the costs of operation and maintenance at the power plant. These include measures such as the exclusion of fine sediments through additional settling basins, reduction of sediment exposure of the turbines through an improved operational plant rule, and an increase the life of the turbines.

A scale model of two additional settling basins was built, as figure 6 illustrates, in order to study the hydraulic feasibility to increase the trapping efficiency of finer particles. The research achieved improved flow distribution and a favourable pattern for the settling basins to trap more sediments.  

The plant’s operation strategy links the generation of the turbines to the silt particle concentration in the suspended sediment. As presented in the figure 7 and shown in the graph of figure 8, if the concentration is below 1,500 ppm, the power plant will operate at full capacity. However, if the concentration is above 3,000 ppm, the power plant will shut down completely. Between 1,500 and 3,000 ppm, the three units are shut down one by one as the concentration rises.

The enhanced maintenance programme proposes changes in maintenance periods and hard coating of the hydromechanical equipment in order to increase turbine efficiency. Just before and after the monsoon, maintenance should be applied to slow down the decrease in efficiency. However, outside the rainy season, the efficiency of the power plant should be higher to produce extra generation. The periods of the maintenance programme are shown in the figure 9.

Ceramic spray coating (R-type), coating of the guide vanes, and the spare of runner sets are the suggestions that have been explored to reduce wear on the turbines and therefore optimise energy generation. However, ceramic coating was not satisfactory. Instead, another type of coating which was used on the guide vanes and applied to one turbine withstood the abrasion. Three sets of turbine parts have been kept to reduce the annual maintenance time to a few hours.

Graphs and figures

‍

‍

‍

‍

‍

‍

‍

‍