Characteristics of ice accretions on blade of the straight-bladed vertical axis wind turbine rotating at low tip speed ratio

Characteristics of ice accretions on blade of the straight-bladed vertical axis wind turbine rotating at low tip speed ratio

Accepted Manuscript Characteristics of ice accretions on blade of the straight-bladed vertical axis wind turbine rotating at low tip speed ratio Yan ...

2MB Sizes 0 Downloads 48 Views

Accepted Manuscript Characteristics of ice accretions on blade of the straight-bladed vertical axis wind turbine rotating at low tip speed ratio

Yan Li, Shaolong Wang, Qindong Liu, Fang Feng, Kotaro Tagawa PII: DOI: Reference:

S0165-232X(17)30417-2 doi: 10.1016/j.coldregions.2017.09.001 COLTEC 2443

To appear in:

Cold Regions Science and Technology

Received date: Revised date: Accepted date:

27 June 2016 18 July 2017 1 September 2017

Please cite this article as: Yan Li, Shaolong Wang, Qindong Liu, Fang Feng, Kotaro Tagawa , Characteristics of ice accretions on blade of the straight-bladed vertical axis wind turbine rotating at low tip speed ratio, Cold Regions Science and Technology (2017), doi: 10.1016/j.coldregions.2017.09.001

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT Characteristics of ice accretions on blade of the straight-bladed vertical axis wind turbine rotating at low tip speed ratio

Yan Li1,*, Shaolong Wang1, Qindong Liu1, Fang Feng2, Kotaro Tagawa3 1-Engineering College, Northeast Agricultural University, Harbin, China

PT

2-College of Sciences, Northeast Agricultural University, Harbin, China

Corresponding author: Tel: +86-451-55191123, Fax: +86-451-55191668

SC

*

RI

3-Faculty of Regional Sciences, Tottori University, Tottori, Japan

Email: [email protected]

NU

Postal Address: No.59, Mucai St. Xiangfang Dist. Harbin, China.

MA

Abstract:

Icing on blade surface of wind turbine greatly affects the performance of wind turbine which

D

becomes a very serious problem for the wind turbine installed in cold and wet regions both for

PT E

larger-scale wind turbine and small-scale ones. The Straight-bladed Vertical Axis Wind Turbine (SB-VAWT) is a kind of lift-type VAWT which is widely used for small-scale wind power

CE

generation. Icing on blade of SB-VAWT will also affect both the starting performance and out power performance. To verify the icing characteristics of blade rotating at low tip speed ratio of the

AC

SB-VAWT during its starting stage, wind tunnel icing experiments have been carried out on a rotor with blade airfoil of NACA0018 in a wind tunnel icing experimental system by using natural low temperature in winter proposed in this study. The icing distributions were recorded by a high speed camera at different tip speed ratios from 0 to 1. The weight of icing on blade was also measured. Furthermore, the icing area ratio and icing rate were calculated and analyzed. The main results show that the ice accretion on a rotating blade of SB-VAWT at low tip speed ratio is quite different from the blade in static condition. The icing distributed on the whole surface of blade and the icing area increases along blade airfoil outline with the increasing of icing time and rotational speed. The 1

ACCEPTED MANUSCRIPT whole weight of icing on all blades of wind turbine increased with the increasing of blade number. Keywords: Wind Energy, Atmospheric Icing, Vertical axis wind turbine, Wind tunnel test, Rotating blade, Tip speed ratio

1. Introduction

PT

With rapid development of large-scale wind turbine, the small-scale off-grid wind turbine has also been concerned in recent years. There are many types of wind turbines. According to the

RI

position of rotor shaft relative to tower or ground, wind turbines can be classified into Horizontal

SC

Axis Wind Turbine (HAWT) and Vertical Axis Wind Turbine (VAWT). With the development of aerodynamic theory of aircraft and being applied successfully for years, the HAWT becomes the

NU

most popular wind turbine in the world, especially in the large scale grid-connected wind power generation. In modern society, with rapid development of small-scale wind turbine for distributed

MA

generation and off-grid wind power market, there is resurgence of interests in VAWT by researchers again which is mainly due to the characteristics of independence from wind direction

PT E

D

comparing with the HAWT [1]. Among all kinds of VAWTs, the Straight-bladed Vertical Axis Wind Turbine (SB-VAWT) has the advantages of simple configuration, low cost and good efficiency. It becomes one of the widely researched VAWT in recent years [2]. Previous studies

CE

show that the SB-VAWT has more advantages when it serves to the small-scale wind turbine which

countryside.

AC

is one of the most important renewable energy resources for people living in urban area and

However, for the wind turbine installed in cold regions, icing on blade is a very serious problem in cold and wet climate which greatly affects output performance of wind turbine and leads to safety accident [3-6]. Two examples of wind turbine icing are shown in Fig.1. Fig.1 (a) shows the icing on HAWT photographed in Sichuan Province of China and Fig.1 (b) shows the icing on SB-VAWT located in Tottori, Japan. Now there are some researches having been done on problem of icing on blade for HAWT [7-11]. However, the characteristics of icing on blade of HAWT are not suitable to

2

ACCEPTED MANUSCRIPT the VAWT, because the working principle and the configuration of rotor are quite different between the HAWT and VAWT. Therefore, the effects of icing on VAWT should be researched. According to the previous researches, only the characteristics of icing on static blade of VAWT have been studied [12], the icing on a rotating blade of VAWT has been seldom reported. Based on the previous researches [1], the power performance of SB-VAWT has two

PT

characteristics that are also two problems in some sense. 1) The aerodynamic performance is not good for the tip speed ratio from 0 to 1, which means that the SB-VAWT has bad starting

RI

performance than HAWT. 2) The maximum power coefficient is lower than that at the high tip

SC

speed of 3 or 4. Therefore, the researches on SB-VAWT recently mainly focus on the two directions, the increasing of output power performance at high tip speed ratio and the improvement of starting

NU

performance at low tip speed ratio. In this research, only the icing effects are focused on the starting

MA

performance of SB-VAWT when wind turbine is used in cold climate regions. The researches on icing effects on rotating blade with high tip speed ratio will be carried out in future work.

D

There are two methods of researching on icing of wind turbine, they are the experimental method

PT E

using icing wind tunnel and simulation method [13, 14]. Due to high demand of equipment and environment, high cost and complex experimental process of icing wind tunnel test, the simulation

CE

is more popular and used as main method to research on icing on blade now. In this paper, a new experimental method is proposed by using natural low temperature in cold region [15]. A common

AC

wind tunnel was reformed by equipping some instruments to supply icing conditions. The experiments have been carried out in winter because the low temperature for icing can be obtained from natural environment which omits the refrigeration system. The main parameters of this reformed icing wind tunnel such as liquid water content (LWC) and temperature have been measured and satisfied the experimental conditions based on verification tests [15]. A model of rotor with different blade numbers of small-scale SB-VAWT was designed and made for icing test. Experimental researches on the characteristics of icing distributions on blade surface of the test rotor have been carried out in this icing wind tunnel. The experiment of icing accretion on the 3

ACCEPTED MANUSCRIPT rotating wind turbine blade has been tested under different tip speed ratios of rotor and discussed. Furthermore, the icing rate and icing area of blade were also calculated and analyzed. 2. Experiments 2.1. Test rotor

PT

The model of rotor designed and made in experiments is shown in Fig.2. The diameter of rotor (D) is 0.44m and the height of blade is 0.25m according to the size of wind tunnel used in this study

RI

of which the size of experimental segment is 0.6m×0.6m×1m. According to the previous researches,

SC

the symmetrical airfoils of NACA series were always selected for research. In this paper the NACA0018 airfoil is used to research on icing on blade. The chord length of blade is 0.125m and

NU

the material is FRP (Fiberglass-Reinforced Polymer). In the experimental study, the wind speed is 4m/s, the diameter of droplet is 40 μm and the surface roughness of blade made in FRP material is

MA

8.4 μm. There were two kinds of blade numbers of rotor: 2 and 3, in order to study the effect of the blade number on icing characteristics. An induction motor was set up under the test rotor outside

PT E

D

the test section. The rotor can be working at a certain rotational speed or tip speed ratio under the control of motor. The tip speed ratio (λ) is defined as shown in Equation 1. The tip speed ratios increases from 0 to 1 by step 0.2. Zero tip speed ratio means that the rotor blade was in the static

R Dn  U 30U

(1)

AC



CE

state. Therefore, this test was only carried out using one blade when λ is zero.

Where n (rpm) is rotational speed of rotor; R (m) is the rotor radius and U (m/s) is the wind speed. 2.2. Experimental System Fig.3 shows the schematic diagram of the experimental system of icing wind tunnel proposed in this research. The wind tunnel is designed and made by Tottori University of Japan which belongs to open type with the size of outlet of 0.6m×0.6m. Water sprayers with different diameters are installed at the outlet of wind tunnel to supply micro water droplet. The tests were carried out in

4

ACCEPTED MANUSCRIPT winter in order to utilize the low temperature outside the lab. The cold air outside was sucked into the wind tunnel and the water was sprayed from the nozzle and mixed with the cold wind at the same time. The flow of water can be adjusted by changing the number of sprayer and be controlled by a flow controller and measured by a flow meter. The three major parameters such as temperature stability, liquid water content and medium volume droplet diameter of the experimental system

PT

have been tested and calibrated in the author`s previous researches. The results show that the major indexes can meet the requirements of the test of icing on blade to a certain degree [15].

RI

Wind tunnel tests were carried out in the winter from December 2014 to January 2015 in

SC

Northeast Agricultural University which is located in the most northeast part of China. The wind speed used in this study was 4m/s. The environmental temperature in wind tunnel ranged from -10℃

NU

to -5℃, which insured that the icing type was glaze ice in this study. Although the temperature

MA

cannot be controlled by this experimental system, the tests were carried out strictly almost under the same temperature condition. The temperature during the test was almost stable during each

D

experiment. Fig.4 shows a temperature curve that the temperature changes in one hour in the test

PT E

segment of wind tunnel. There were two different densities of average liquid water content (LWC) used in this study, 1.16g/m3 and 2.32g/m3, which were the averaged values calculated based on the

CE

wind speed and water flow discharges. The averaged diameter of water droplet (dm) is about 40μm in this study which can be calculated by the equations below [14]: 1(1 / 3)

dn

AC

d m  3.21We

(2)

We 

wV 2d n w

(3)

V 

Qw d n2 / 4

(4)

Where, We is the Weber Number, dn is the diameter of sprayer,  w is water density, V is water speed,  w is surface tension and Qw is water flow discharge.

5

ACCEPTED MANUSCRIPT To capture the distribution of icing on blade surface in rotational state, a high speed camera has been used in this study is shown in Fig.5. The camera type is Phantom v5.1 made by The Vision Research Inc (VRI), USA. The resolution ratio is 1024 x 1024 pixels. The camera was connected with a computer, and the photos can be sent to the computer by the Phantom Camera control software shown in Fig.5. The ice distributions and icing area were measured by image processing

PT

software.

RI

3. Experimental results

SC

3.1 Icing Accretion

3.1.1 Icing accretions on rotating blade along span direction

NU

A result of the ice distribution on blade along the span direction is shown in Fig.6. From that the

MA

ice thickness at any position along span direction was almost the same. Therefore, the icing shape of airfoil section photographed by high speed camera from the top of the rotor can represent the

D

characteristics of icing on blade airfoil.

PT E

3.1.2 Icing accretions on static blade

The distributions of icing on blade surfaces at different attack angles in static condition to

CE

compare with the blade in rotational condition are shown in Fig.7. The characteristics of ice accretion on static blade depend on the attack angles greatly as a whole. At 0 degree, the icing only

AC

occurs at the part of leading edge and grows towards to the position with maximum thickness of blade. For other angles, the ice accretion increases from leading edge towards to trailing edge with the increasing of attack angle and distributes on the whole surface facing to wind. On the contrary, there is almost no icing on the other side of blade. In another word, the icing cannot cover the whole surface when the blade is in static condition. For more details about the characteristics of icing on static blade, some previous works can be learned from the reference [12]. 3.1.3 Icing accretions on rotating blade

6

ACCEPTED MANUSCRIPT Fig.8 ~ Fig.11 show photographs of icing accretion each five minute interval during 30 minutes on rotating blades of the test rotors with different blade numbers under different icing conditions at the tip speed ratios of 0.2 and 1. Furthermore, the icing shape was induced into image soft and the icing outline was obtained also shown in these Figures. According to the results, the icing occurs on the whole blade surface for all test conditions, which

PT

is different from the characteristics of icing on blade in static condition. At the beginning of icing in first 5 minutes, the ice accretion is a thick layer grown along blade airfoil outline. With increasing

RI

of time, the ice accretion is grown along the previous layer. In addition to the leading edge and

SC

trailing edge of the blade, the shape of the airfoil is not significantly changed. Therefore, it can be concluded that the wind turbine can still maintain a better performance in the initial stage of icing to

NU

a certain extent. With increasing of freezing time, the shape of trailing edge is no longer a sharpen

MA

point. It becomes a circular arc shape. It can be also found that the ice thickness of the leading edge is obviously thicker than that of the other parts. However, the change regulation of this result is

D

different with variation of blade number of rotor, the rotational speed of rotor and the liquate water

PT E

content. Fig.12 shows the schematic diagram of local speed around single blade at different attack angles which can explain this experimental result. The changes of relative wind velocity (W) at different rotating angles are presented. The relative wind velocity of the blades at different

CE

rotational angles changed because of the existence of rotational speed. As a result this leads to the

AC

increase of probability accepting water droplets and icing at where near the leading edge and the trailing edge facing the windward side than other parts of blade surface. At the same time, the trend of icing accretion towards to the trailing edge of blade because of the centrifugal force generated by rotation. However, the relative wind velocity has not been changed largely at relatively low tip speed ratio such as λ=0.2, which makes that the trend is not obvious than the case of higher tip speed ratio (λ=1). It can be predicted that when the SB-VAWT is working at more higher tip speed ratios such as λ=2 or 3 in the same icing conditions, the ice accretions will become more serious at the leading edge and trailing edge of blade, which will make the outline of iced blade change into

7

ACCEPTED MANUSCRIPT the rectangular shape and no longer maintain the airfoil shape. The blade will lose the aerodynamic characteristics which leads to dynamic stall and stop running finally. In addition, the difference of LWC and blade number also has the effects on ice accretions on blade surface. The larger LWC will make the icing on blade become thicker than the case of lower LWC. 3.2 Icing area

PT

To analyse the characteristics of ice accretions on blade deeply, two coefficients about icing area

RI

are defined. They are net icing area rate (NIAR: δ) and total icing area increasing rate (TIAR: β) respectively. The definition of icing area is shown in Fig.13. The calculations of NIAR and TIAR

SC

are shown in Tab.1 and Tab.2. The NIAR indicates the net increased icing area each 5 minutes

NU

interval. The TIAR means that all increased icing area until some certain time. Fig.14 ~ Fig.17 show the NIAR and TIAR of the test rotors with different blade numbers under different icing

MA

conditions at the tip speed ratios from 0.2 to 1.

According to Fig.14 and Fig.15, the NIAR increases rapidly in the first 5 minutes for all the cases.

D

However, after a rapid falls during the next 5 minutes the NIAR tends to be almost steady and

PT E

changes little in a small range. The change range is different with variation of LWC and blade number. The result can be reasoned that at the beginning of icing process the super-cold water

CE

droplet collides and adheres to the cold blade surface initially with no icing and then becomes icing rapidly. In this stage the super-cold water droplet contact the cold solid surface directly. The

AC

temperature between them is quite different and the heat transfer is fast, which makes the icing rate become the fastest. When there are some certain ice accretions the contact becomes to be between the super-cold water and icing only and then the icing area increasing rate trends to be kept changing in a small range. However, it can be found that the NIAR of λ=1 is lower than NIAR of some other tip speed ratios. Although the reasons for this phenomenon are not fully understood for us, the centrifugal force can be regarded as one of the main reasons. The centrifugal force increases with increasing of rotation speed. When the super cooled droplets fall on the surface of blade, it couldn’t freeze instantly and part of them are thrown out of the surface by the centrifugal force. 8

ACCEPTED MANUSCRIPT This will affect icing and ice accretion on blade in some degree. Because that we did not measure the centrifugal force in the test, we cannot explain clearly how the centrifugal force affect the icing characteristics on rotating blade. Therefore, the test results obtained are only suitable for the SBVAWT at low tip speed ratio; the effects on the rotor with high rotational speed may be different. According to Fig.16 and Fig.17, the characteristics of TIAR are different from the NIARs’. The

PT

TIAR has a linear growth with the increasing of time for all cases. The increasing gradient is different for the rotors at different tip speed ratios and LWC conditions. There is also some

RI

difference of TIAR for the rotor with different blade number. For the rotor with three blades, the

SC

total icing area becomes more than the airfoil section area at λ=1 and LWC=2.32 g/m3 in 30 minutes. It can be predicted that the total icing area will be kept increasing and becomes more than

NU

the airfoil section area if the icing time is enough long.

MA

3.3 Icing weight

The weights of icing both on a single blade and all the blades of the test rotor totally in 30

D

minutes are shown in Fig.18. Generally speaking, the whole icing weight adhered on all blades of

PT E

the rotor with three blades is more than the rotor with two blades. It concludes that the rotor will capture more super-cold water and become icing with increasing of numbers of blades. This trend is

CE

more pronounced for high LWC condition. However, for the icing weight on a single blade, the trend is just the opposite only except for the case of tip speed ratio λ=1. This can be explained as

AC

that the more blade number the rotor has along the circumferential direction, the greater the blades interfere each other, which leads to the weight of icing on a single blade of the three blades rotor is lighter than that of the two blade rotors although the total weight of icing on the three blades rotor is heavier. To further analyze how much the super-cold water from wind tunnel sprayer, a coefficient called icing rate (γ) is defined as below:

9

ACCEPTED MANUSCRIPT  

M A1 Q t A2

 100%

(5)

Where, M is the weight of icing, Q is the mass of flow discharge of cold water and A1 is the swept area of rotor (A1=H×D), A2 is the area of experimental section, A2=0.6m×0.6m, t is testing time.

PT

The icing rates for a single blade and all the blades of test rotor totally in 30 minutes are shown in

RI

Fig.19. The icing rate for the rotor with two blades has a maximum value at the tip speed ratio from

SC

0.6 to 0.8. However, for the rotor with three blades, the icing rate rises with the increasing of tip speed ratio under the test condition in this study. It indicates that the icing ability for a single blade

NU

at a certain rotational speed has a limit or a maximum value. The number of blade will greatly affect the rotor’s icing ability and the tip speed ratio at which the icing rate can reach to maximum value.

MA

Also, the LWC is an important factor affecting the icing rate. Furthermore, there is one point should be noted that there are some fluctuations of weights of

PT E

D

icing measured and icing rates calculated. There are two influential factors. One is that there is still some inhomogeneity for the icing on the whole blade surface along the span direction which makes the icing weight change a little. The other is that the test experimental temperature has a slight

CE

difference for the tests. However, it does not affect the overall characteristics of ice accretion on

AC

blade. 3.4 Discussions

3.4.1 Icing occurrence and accretion There are only two states for the wind turbines, rotating and non-rotating. For the wind turbine in static condition, it may be suffered the wind with super-cold water from various directions. Therefore, the icing can occurs anywhere of the blade surface towards to wind direction as shown in Fig.7. If the blade airfoil wind turbine used is a kind of symmetrical airfoils such as NACA0018 test in this study, the ice accretion on one side surface towards to wind direction will change the airfoil 10

ACCEPTED MANUSCRIPT section from symmetrical shape to asymmetrical shape, which leads to a complete change in aerodynamic characteristics of the airfoil. Therefore, the wind turbine with iced blades will not begin to rotate. For the wind turbine with asymmetrical blade airfoil, the icing on blade will increase the asymmetry of airfoil gradually with increasing of ice accretion. However, parts of the aerodynamic performance may be kept for iced blades at the initial stage of the icing occurrence.

PT

Therefore, the wind turbine can still have the ability to start operations until that the ice accretion makes the aerodynamics characteristics of blade airfoil completely lost. When the blade with

RI

symmetrical airfoil is iced in the rotating process like the cases researched in this study, the ability

SC

of keeping rotation for wind turbine with iced blade is quite different under different icing condition and for different wind turbine structures including blade number, rotational speed, etc. Because the

NU

icing on wind turbine with asymmetrical airfoil was not test in this study, this part should be further

MA

researched by asymmetrical airfoils developed and used by SB-VAWT recently. Furthermore, the effects of ice accretions on rotational blade on aerodynamic characteristics should be test or

D

evaluated in the future study.

PT E

3.4.2 Icing pattern

The pattern of icing created by the wind tunnel icing experimental system in this study belongs to

CE

glaze ice. This is not only concluded by the temperatures (-10 ℃ to -5 ℃) measured in the tests but also observed from all the photographs taken by high speed camera. According to ISO 12494 [17],

AC

the atmosphere icing can be catalogued into three patterns: Rime Ice, Glaze Ice and Mixed Ice. Many previous researches, especially for the numerical simulation researches, focused on mainly rime ice on wind turbine blade. The study on mechanism and characteristics of glaze ice on wind turbine blade researched mainly by icing wind tunnel tests is rare. It is concluded that the effects of glaze icing on blade on the rotation characteristic and the structural strength of the wind turbine is much more than that of the rime icing on blade because of the high density of glaze ice than rime ice. Therefore, the results obtained in this test can be seen as the worst effects of icing on rotating blade of VAWT in some sense. 11

ACCEPTED MANUSCRIPT 3.4.3 Icing on VAWT and HAWT Although the object of this study is researching the icing on rotating blade for vertical axis wind turbine, the test methods and the principle of analysis can be also applied to the horizontal axis wind turbine. The main difference between the two types of wind turbine is the rotation mode. For the SB-VAWT, the rotating blades will make the rotor become a cylindrical shape. In another word, it’s

PT

a three-dimensional space. Therefore, the blades will have more chance for getting more super-cold water and icing during the wind flowing across the rotor especially for the condition of high

RI

rotational speed. However, for the HAWT, the swept area of rotating blade is a plane. In another

SC

word, it’s a two-dimensional space, in some sense, which makes that the blades only have one surface to capture super-cold water and become icing when the wind flow across the rotor.

NU

Furthermore, the characteristics of ice accretion on rotating blade for the two types of wind turbine

MA

will also be different greatly because the rotation direction of blade against wind direction is different completely. The analysis method shown in Fig.12 used for SB-VAWT in this study can

D

also be a reference benefiting for the study about ice accretion on rotating blade of HAWT.

PT E

4. Conclusions

Under the conditions of this study, the main conclusions can be summarized as below:

CE

1) The ice accretion on a blade rotating at low tip speed ratio of the Straight-bladed Vertical Axis

AC

Wind Turbine during starting stage is quite different from the blade in static condition. The ice distributed on the whole surface of blade and the icing area increased along with blade airfoil outline, which is the most basic and important characteristics of ice accretion on rotating blade operating under the low tip speed ratios smaller than one. 2) The blade airfoil changes from symmetry to asymmetry with the increasing of icing time and rotational speed. The shape of trailing edge changes from a sharpen tip to arc shape and the icing near leading edge becomes significantly thicker than other parts of blade. One of the main reasons is the changes of local relative wind velocity at blade under low tip speeds ratios.

12

ACCEPTED MANUSCRIPT 3) The blade numbers affect the weight of icing on a single blade and the whole rotor. The whole icing weight adhered on all blades of the rotor increased with the increasing of blade number. On the contrary, the result of weight of icing on a single blade is just opposite. This is the characteristics of ice accretion on rotating blade of SB-VAWT operating under the low tip speed ratios smaller than one.

PT

Acknowledgments

RI

This research is sponsored by the Project 51576037 and 10702015 supported by National Natural

SC

Science Foundation of China (NSFC). The authors give thanks to their supporters. References

NU

[1] Paraschivoiu, I. Wind Turbine Design.Polytechnic international Press 2002.

MA

[2] Islam M, David S, Ting K, et al. Aerodynamic models for Darrieus-type straight-bladed vertical axis wind turbines.Renewable and Sustainable Energy Reviews2008; 12: 1087-1109.

D

[3] T.Laakso, L. Talhaug, K.Vindteknik, G.Ronsten, R.Horbaty, I.Baring-Gould, A.Lacroix, E

PT E

Peltola. Wind energy projects in cold climates (First edition).Executive committee of the International Energy Conversion Systems2005; 1-36.

CE

[4] Nalili. N, A. Edrisy, R. Carriveau.A review of surface engineering issues critical to wind turbine performance.Renewable and sustainable energy reviews 2009; 31:428-438.

AC

[5] Matthew C. Homola, Muhammad S. Virk, Per J. Nicklasson, Per A. Sundsb. Performance losses due to ice accretion for a 5 MW wind turbine.Wind Energy 2012; 15: 379-389. [6] M. Etemaddar, M. O. L. Hansen, T. Moan. Wind turbine aerodynamic response under atmospheric icing conditions. Wind Energy 2014; 17: 241-265. [7] Bose, N. Icing on a small horizontal axis wind turbine – Part 1: Glaze ice profiles. Journal of Wind Engineering and Industrial Aerodynamics 1992; 45: 75-85. [8] Bose, N. Icing on a small horizontal axis wind turbine – Part 2: Three dimensional ice and wet snow formations. Journal of Wind Engineering and Industrial Aerodynamics 1992; 45: 87-96. 13

ACCEPTED MANUSCRIPT [9] Homola Matthew , Wallenius Tomas , Makkonen Lasse , Nicklasson Per , Sundsb Per. Turbine size and temperature dependence of icing on wind turbine blades.Wind Engineering2010; 34:615-627. [10] Andrea G. Kraj, Eric L. Bibeau. Measurement method and results of ice adhesion force on the curved surface of a wind turbine blade. Renewable energy2010; 35: 741-746.

PT

[11] Andrea G. Kraj, Eric L. Bibeau.Phases of icing on wind turbine blades characterized by ice

RI

accumulation.Renewable energy2010; 35:966-972.

[12] Yan Li, Kotaro Tagawa, Fang Feng, Qiang Li, Qingbin He. A wind tunnel experimental study

SC

of icing on wind turbine blade airfoil.Energy Conversion and Management 2014; 85: 591-595.

NU

[13] Yiqiang Han, Jose Palacios, Sven Schmitz. Scaled ice accretion experiments on a rotating wind turbine blade.Journal of Wind Engineering and Industrial Aerodynamics 2012; 109: 55-67.

MA

[14] F Villalpando,M Reggio,A Ilinca. Numerical study of flow around iced wind turbine airfoil. Engineering Applications of Computational Fluid Mechanics 2012; 6: 39-45.

PT E

D

[15] Li Y, Wang S L, Zheng Y F, Liu Q D, Feng F, Tagawa K. Design of wind tunnel experiment system for wind turbine icing by using natural low temperature.Journal of Experiments in Fluid Mechanics 2016; 30: 54-58, 66.

CE

[16] Faeth G M, Hsiang L P, Wu P K. Structure and breakup properties of sprays. International

AC

Journal of Multiphase Flow 1995; 21: 99-127. [17] ISO 12494. Atmospheric icing of structures.1st ed. Geneva; 2001-08-15.

14

PT

ACCEPTED MANUSCRIPT

(b) VAWT

RI

(a) HAWT

PT E

D

MA

NU

SC

Fig. 1 Examples of icing on wind turbine blades

CE

Fig. 2 Schematic diagram of model rotor

Water Sprayer

AC

Test rotor

Wind tunnel

Cold air outside Water and air outlet

Fig. 3 Schematic diagram of wind tunnel icing experimental system

15

ACCEPTED MANUSCRIPT

Temperature(degree)

1 0 Test time: 2015.1.23 In one hour from wind tunnel operation starting

-1 -2 -3 -4 -5

PT

-6 -7 20 30 40 Time(minute)

50

60

RI

10

SC

0

PT E

D

MA

NU

Fig. 4 Temperature changes in one hour in the test section of wind tunnel

AC

CE

Fig. 5 High speed camera used and the image processing software

Fig. 6 The icing distribution along blade span direction

16

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

Fig. 7 Icing distributions on blade with NACA0018 airfoil in static condition at U=4m/s

17

ACCEPTED MANUSCRIPT

10min

15min

20min

25min

PT

5min

0min-30min

RI

30min

10min

15min

20min

25min

PT E

D

5min

MA

NU

SC

(a) λ=0.2

30min

0min-30min (b) λ=1

AC

CE

Fig. 8 Ice accretions on blade of two blades rotor (averaged LWC=1.16g/m3)

18

ACCEPTED MANUSCRIPT

10min

15min

20min

25min

PT

5min

30min

0min-30min

10min

15min

20min

25min

PT E

30min

D

MA

5min

NU

SC

RI

(a) λ=0.2

0min-30min (c) λ=1

AC

CE

Fig. 9 Ice accretions on blade of three blades rotor (averaged LWC=1.16g/m3)

19

ACCEPTED MANUSCRIPT

10min

15min

20min

25min

PT

5min

30min

0min-30min

PT E

30min

15min

MA

10min

D

5min

NU

SC

RI

(a) λ=0.2

20min

25min

0min-30min (d) λ=1

AC

CE

Fig. 10 Ice accretions on blade of two blades rotor (averaged LWC=2.32g/m3)

20

ACCEPTED MANUSCRIPT

15min

20min

25min

PT

10min

5min

0min-30min

RI

30min

10min

15min

20min

25min

PT E

D

5min

MA

NU

SC

(a) λ=0.2

0min-30min

30min

(e) λ=1

AC

CE

Fig. 11 Ice accretions on blade of three blades rotor (averaged LWC=2.32g/m3)

21

AC

CE

PT E

D

MA

NU

(a) λ=0.2

SC

RI

PT

ACCEPTED MANUSCRIPT

(f) λ=1 Fig. 12 Schematic diagram of local speed around single blade at different attack angles

22

RI

Fig. 13 Definition of icing area

PT

ACCEPTED MANUSCRIPT

SC

Table 1 Net icing area rate around blade surface T

Net Icing Area Rate (NIAR)

0

5min

1

10min

2

…..

……

…..

30min

6

S6  100% S

PT E

D

MA

NU

0min

0%

S1  100% S S2 100% S

Total Icing Area Rate (TIAR)

AC

T

CE

Table 2 Total Icing Area Rate around blade surface

0min

0

0%

5min

1

10min

2

S1  100% S S1  S 2  100% S

…..

……

…..

30min

6

S1  S 2  S3  S 4  S5  S 6 100% S

23

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

(a) averaged LWC=1.16g/m3

(b) averaged LWC=2.32g/m3 Fig. 14 Net icing area ratio of two blades rotor

24

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

(a) averaged LWC=1.16g/m3

(b) averaged LWC=2.32g/m3 Fig. 15 Net icing area ratio of three blades rotor

25

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

(a) averaged LWC=1.16g/m3

(b) averaged LWC=2.32g/m3 Fig. 16 Total icing area ratio of two blades rotor

26

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

(a) averaged LWC=1.16g/m3

(b) averaged LWC=2.32g/m3 Fig. 17 Total icing area ratio of three blades rotor

27

MA

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

(a) all blades

(b) single blade Fig. 18 Icing weight 28

NU

SC

RI

PT

ACCEPTED MANUSCRIPT

AC

CE

PT E

D

MA

(a) all blades

(b) single blade Fig. 19 Icing rate

29

ACCEPTED MANUSCRIPT Highlights 1.

We experimentally investigated ice accretion on ratating blade of SB-VAWT operating in starting stage.

2.

The icing characteristics on rotating blade under different tip speed ratios from 0 to 1 and icing conditions were obtained. The icing area rate and icing weight rate were calculated and analysed.

4.

The difference of ice characteristics was compared between the rotating blade and static blade.

AC

CE

PT E

D

MA

NU

SC

RI

PT

3.

30