Jhimpir is characterized as an arid desertic region having dry land with little agricultural activities mainly in its southern part and located at 25° 01' 32'' N and 68° 00' 47" Approximately 135 Km in the northeast of Karachi city. The entire Jhimpir region has low precipitation, little to no vegetation and mostly rocky & gravely soil texture like rest of the coastal region of Pakistan (Javid et al., 2019; ADB, 2016; Ghalib et al., 2014). Hussainet al. (2005) observed that in spite of having frequent dust events, the frequency of dust storms in the region is quite low despite the arid nature of the entire Sindh province. This may be due to the fact that the whole geomorphic region does not act as single point source and multiple erodible point sources cause the dust production (Reynolds et al., 2007; Eckardt and Kuring, 2005; Gillette, 1999; Gillete et al., 1997). Walker et al. (2009) analyzed the dust sources in southwest Asia and did not pointed out any major erodible point source in Sindh region. Moreover, the region has the maximum bench mark wind speed of 8.8 m/s at height of 30 m (Mirza et al., 2010); whereas, in order to report the dust storm, the wind speed must be equal to or exceed than 22 Knots or 11.32 m/s (Hussain et al., 2005), and not high enough to produce dust storms. Thus, locally produced dust events are mostly dominated by drifting, blowing and suspending dust. The observed dust storms in the study region are usually not local to the area and originally initiate from the border of Iran, Afghanistan and Pakistan. 34% of these guest storms has been observed in summer season (i.e. form May till July), followed by 28% in winter season (i.e. from November till January), 21% in autumn (i.e. from August till October) and 17% in spring season (i.e. from February till April) during the period of 1999 till 2018. Figure 4infers that highest number of dust storms occurred in the month of June and December; while, average numbers of dust storms have been observed in the months of March, August, July, September and November respectively.
Beside these foreign dust storms, local dust events mostly comprise of blowing dust, drifting dust & suspended dust which can be observed all around the year. These local dust events are usually influenced by two factors, namely (a) high wind speeds, and (b) very low precipitation. The high and turbulent wind speeds near the land can causes more dust to blow from dry erodible terrain. As shown in Fig. 2, the observed wind speeds in months of May to August and November to December are relatively high, which makes these months more prone and susceptible for the occurrence of local dust events in the study area.
On the other hand, during the rainy season the moisture content of the land increases causing dust to stick to the land and offers resistance in easily blowing off of the dust. High mean monthly precipitation has also been observed during July to September with its peak in August for period of 2009 to 2020 (Table 1). However, in southern region of Pakistan comprising of the study area, the rainfall is commonly preceded by thunderstorms and extremely high wind and after the rainfall when land dries, the dust & sand on ground becomes extremely erodible and blows off quite easily, resulting in increased dust activity in this season. Rezazadeh et al. (2013) recorded the highest number of dust events in the month of July.
Various other research studies also supported the fact that the dust activities in Pakistan are comparatively highest during the summer season (Sharif et al., 2015, Rezazadeh et al., 2013; Alam et al., 2012; Hussain et al., 2005). Furthermore, Alam et al. (2012) studied aerosol optical and radiative properties and indicated dominance of coarse particles in summer and fine particles in winter season in and around Karachi (nearby to the study area) and Lahore.
The Jhimpir region is humid in nature as a part of being the coastal region of Pakistan and having freshwater bodies in the vicinity such as Haleji lake, Keenjhar lake and Hadero lake. Average range of humidity is between 30–50%. Table 1 shows that May to October are the humid months while months of July to September are extremely humid due to monsoon season. The increased moisture content of the air causes the finer dust particles or other suspended matter in the air to stick and accumulate on any surface, like different parts of a wind turbine in our case, causing an increase in the surface roughness as well as in the mass of turbine blade. The wind turbine efficiency is highly dependent on the blade design, mass and its smoothness, which means that any change in these parameters may affect the electrical output power of the wind turbine.
Therefore, during the investigation of the impact of dust events on the electrical output and the efficiency of a wind turbine within the Jhimpir region, several findings have been made. However, as shown in Table 2, the acquired electrical output data of turbines has certain unexplained discrepancy for the months of September, November & December probably due to data accumulation error of windfarm’s software and therefore these months have not been considered during the final calculation of percentage error between ideal (expected/theoretical) power and actual power.
Table 2 illustrates an average positive percentage error between theoretical and actual power (𝝃TA) & percentage error between expected and actual power (𝝃EA) of 5–7% in month of January as it has already been discussed earlier that months of November and December are prone to occurrence of local as well as foreign dust events which causes the dust to blow and accumulate on the turbine blade resulting in increased percentage error in January. As months of February till April are dry months therefore the accumulated dust removes away from the blade and percentage error decreases. The percentage error began to increase again in month of May and continued to increase in months followed i.e. till October. Although, the monsoon season also occur in months of July to September, however as explained above, the dust events are quite common in monsoon season before and after the rainfall which resulted in the increased percentage error.The negative percentage error phenomenon has been observed during months of February till June which represents that the actual produced electrical output is higher than the ideal output, which is practically impossible. Though, it occurred due to the fact that anemometer used for wind speed measurement are mounted on the nacelle at the back of the turbine blade which results in underestimated wind speed values and in turn theoretical electrical output values.
In such a dusty environment, accumulation of dust or erosion of turbine blade surface is observed commonly. An experimental investigation performed by Khalfallah and Koliub (2007) to study the effect of dust on the output of the wind turbine showed that accumulation of dust on turbine blade results in degradation of the turbine output. Study also showed that dust size as well as area covered by the dust increases linearly with the passage of time, and depend on various factors such as wind speeds, wind direction, dust storm frequency, and topography of the site and roughness of the blade. Moreover, leading edge of the wind turbine blade collect more dust during rotation as compare to the trailing edge as evident from Fig. 5.
Surface erosion is the phenomenon which is observed in such conditions and highly dependent on the particle Stokes number which is the ratio of particle relaxation time to the flow characteristic time and depends upon diameter of the particle, density of the particle and inflow velocity of the particle. A study conducted by Li et al. (2018) showed that there exist a critical range of particle Stokes number above which blade erodes while below the critical range nothing happens. Furthermore, the erosion areas of the airfoil and erosion rates essentially remain same for the same Stokes number and angle of attack (AOA), even if the controlling parameters are different. However, the airfoil shape has great effects on its erosion, i.e. the larger the curvature of the leading edge is, the more likely the airfoil is to undergo erosion.
In this context, there may be two reasons for the degradation of the turbine electrical output:
1. Increased surface roughness of blade due to accumulation or erosion.
2. Increase mass of the turbine blade due to deposited dust.
Surface roughness is a factor that affects the output of the turbine and can be defined as extension of surface into the fluid and may occur due to dust accumulation or surface erosion. This factor essentially changes the airfoil shape and characteristics. Rezig et al. (2019) performed experimental investigation and deduced that eroded airfoil attains lower lift coefficient (CL) and higher drag coefficient (CD) as compared to the clean foil.
Surface roughness increases the interaction between the fluid and surface which results in increased friction or drag force and drag coefficient (CD); further, the irregularities in the surface layer causes the disturbance in the boundary layer which may results in an early transition from laminar to turbulent flow and decrease in lift force & lift coefficient (CL). As efficiency of the turbine blade is proportional to lift force and inversely proportional to drag force, the phenomenon of increased surface roughness causes a significant decrease in efficiency of wind turbine. However, Ren and Ou (2009) showed that decrement of CL and increment of CD occurs rapidly till a certain critical roughness height and after this value the changes occur very slowly, as the whole airfoil becomes turbulent boundary and increase in roughness remains no longer evident.
For an area with high dust concentration in air, a thin coat of the dust formed over any surface including turbine blade which causes the mass of the blade to increase from usual. As a result of this, kinetic energy of the wind does not efficiently convert into the torque of the blade which results in the decreased electrical output.
Since the specification of turbine blade under study is not available, a similar turbine blade LM 48.8 is used for estimation of the increased blade mass. Aforementioned blade has blade length of 48.7 m, surface area of 119 m2 and blade mass of 9,950 Kg without blade root bolts (PHA, 2009). Table 3 shows the estimated increase in blade mass after 3, 6, 9 and 12 months due to dust accumulation in the study area. It can be seen that the blade mass is increasing linearly with the passage of time and approximately 86.87 Kg mass is increased for single blade while for overall 3 blades 260.61 Kg mass increased after 12 months by assuming 2 g/m2/day dust deposition with moister content. This increased mass resulted in increased rotational inertia which causes decreased angular acceleration. In other words, the wind turbine blade will move slowly for the same wind energy and thus produce less electrical output.
During these calculations it has been assumed that dust accumulate as a uniform layer on the turbine which is not practically correct as already discussed above that leading edge of the turbine blade collects more dust. However, rotational inertia not only depends on mass but also on the position of the mass i.e. the mass has greater impact the farther it is placed from the central position. Although, the whole blade may not fully and uniformly cover with dust, yet the dust accumulation on leading edge of the blade, which is essentially the farthest position from the central hub, causes the rotational inertia increased so much so to result in decreased angular acceleration of turbine blade and turbine efficiency.
Table 3
Dust deposition on turbine after 3, 6, 9 and 12 months.
Dust density deposition
|
2g/m2/day
|
Dust deposition on 119 m2
|
0.238 Kg/ day
|
Dust deposition after 3 months
|
21.42 Kg
|
Dust deposition after 6 months
|
43.08 Kg
|
Dust deposition after 9 months
|
64.97 Kg
|
Dust deposition after 12 months
|
86.87 Kg
|