Wind Feasibility and Optimization Study in East Tennessee

8/13/2019 Wind Feasibility and Optimization Study in East Tennessee

1/19

Wind Power Feasibility and Optimization Study inEast Tennessee

In submission for the Chancellors Honors Program Senior Thesis project

Megan SchuttAerospace Engineering

Mechanical, Aerospace, and Biomedical Engineering Department

University of Tennessee, Knoxville, Tennessee


2/19

Wind Power Feasibility and Optimization Study in

East Tennessee

Megan Schutt

Aerospace EngineeringMechanical, Aerospace, and Biomedical Engineering Department

University of Tennessee, Knoxville, Tennessee

Wind energys application in the East Tennessee and West North Carolina

area was addressed, as well as regions with similar topography and

climatology. A Geographic Information System (GIS) study was developed

using data from the Wind Energy Resource Atlas. These regions relatively

small wind energy sources are optimized through a tailored airfoil and the

investigation of less invasive turbine technologies. In this way, smaller and

more efficient farms can be constructed, lessening the impact on wildlife and

the surrounding community.

Nomenclature

= density

P = power

= mass flow rate

v = velocity

A = rotor area

dE/dt = kinetic energy change

L = lift force

D = drag forceCL = coefficient of lift

CD = coefficient of drag

CP = coefficient of pressure

L/D = lift to drag ratio

= angle of attack

1. IntroductionIn order to lessen dependence on fossil fuels, an

alternative energy supply study remains

necessary. Especially in regions overcome withcurrent land abuse through such resource

gleaning as mountaintop removal, traditional

coal mining and hydraulic fracturing,

unconventional methods may need to be

optimized. For wind farms to become feasible in

mountainous regions such as those in the East

Tennessee and Western North Carolina

Appalachian region, different power

maximization technologies may be

advantageous. This study will attempt to

localize regions prime for wind farm

development through Geography Information

Systems (GIS) as well as to create a solution for

wind turbine technology suitable for such areas

with customized turbine type selection and

unique airfoil design.

2. Theory2.1Geographic Study

First, ideal locations must be scouted according

to average wind speed and temporal variability,

and practical topography. A sound criticism of

wind farms as an anchor energy resource is the

inconsistency in availability. This cannot be

completely alleviated without large gains in

energy storage capabilities, but an expectedsupply can be denoted with analysis of data in

regards to historical patterns throughout

different parts of the year. In fact, some studies

have shown that diurnal wind cycling is actually

in sync with power demand, with higher wind

speeds during the early evening, and tapering

off until the early hours of the morning.1Using


3/19

the Wind Resource Energy Atlas from the US

Department of Energy, wind categories can be

mapped for any state in the country.2 In Table

1, the categories are listed in terms of both

average wind speed, and resultant wind power

density available. In addition, Figure 1 shows a

graphical approximation of wind resource

capability according to geographical features.

Figure 1: NREL: RREDC

Wind

Power

Class*

10 m (33 ft) 50 m (164 ft)

Wind Power Density

(W/m2)

Speed(b)

m/s (mph)Wind Power Density

(W/m2)

Speed(b)

m/s (mph)

10 0 0

100 4.4 (9.8) 200 5.6 (12.5)2

150 5.1 (11.5) 300 6.4 (14.3)

3

200 5.6 (12.5) 400 7.0 (15.7)

4

250 6.0 (13.4) 500 7.5 (16.8)

5

300 6.4 (14.3) 600 8.0 (17.9)

6

400 7.0 (15.7) 800 8.8 (19.7)

71000 9.4 (21.1) 2000 11.9 (26.6)

Table 1: NREL Wind Atlas Categories with associated power densities and wind speeds at heights of 10 m and 50 m

Maximum power from a wind turbine, or any

turbine for that matter, can be estimated using

the mass flow rate across the given rotor area.

Area can be controlled by increasing or

decreasing the rotor diameter, but locations

must be investigated to maximize available

wind, as power is most sensitive to changes in

velocity. Ideally, wind speed would be taken at

each potential turbine point, and a power

distribution function would be modeled off of

these individual values.3

While average wind speed is a good

resource for determining the ultimate power

output available and technology best suited in a

particular area, a more detailed analysis is

necessary to define the expected supply on aday-to-day basis. In addition, directional

information is important to ensure a laminar,

one directional flow that is most useful to

horizontal-axis wind turbines (HAWTs). In

locations where wind speed is low and/or

direction is most variable, vertical-axis wind

turbines (VAWTs) would perhaps be utilized.4


4/19

NREL also offers a theory that exposed

mountain ridges and summits offer more

promising wind resources than perhaps even

average wind speed as a result of the relative

topography and orientation to prevailing winds.

A Venturi speed-up effect happens as the

wind goes over these ridges and the flows are

compressed.5 This instance has already been

utilized by TVA with the one wind energy

generation plant in East Tennessee on Buffalo

Mountain (see Figure 2 below).

Figure 2: Buffalo Mountain Wind Farm, picture courtesy

of American Council on Renewable Energy.

GIS has long been used along with

various data mining sources to graphically

analyze the wind available in particular regionsacross the world.6,7 The main benefit behind

this method is to utilize existing infrastructure

and historical data. After any necessary

modifications in format are made, the same

data used for the sciences of meteorology and

environmental studies can be employed in

modern studies of energy implementation such

as these.

2.2Airfoil DesignGeographic wind availability may indicate

potential for wind energy generation, but the

technology for sourcing this energy can severely

curtail ultimate power generation if not

developed in an optimum way. Improvements

in airfoil design can be a method in which

efficiency gains maximize the power generation

of wind farms.

In design of airfoils for use in HAWTs,

the NACA 63-4XX and NACA 63-6XX series have

already been developed. There have also been

improvements in by the DU xx-W-xxx series, the

S8xx series, and the Riso-A1-xxx series. General

design goals include high maximum lift to drag

ratio, high maximum lift, and insensitivity to

leading edge roughness,8 as HAWTs are

generally lift based mechanisms. VAWTs, such

as Darrieus and Savonius Rotor designs, are

generally drag based9, so their airfoil design

characteristics will have different goals.

Lift and L/D maximization goals

contribute toward increasing power generation,

in the same way that the thrust available in

aircraft engines increases available power,

albeit with a dependency on velocity. The other

design goal of insensitivity to roughness isnecessary, especially in wind turbine usage, to

eliminate effects of particulate matter on

aerodynamic performance. Roughness can lead

to quicker transition of laminar to turbulent

flow if the airfoil is not sufficiently insensitive.

In order to increase insensitivity to

roughness, several techniques can be

employed, that can both negatively and

positively affect the other design goals of lift

and lift/drag maximization. One means of

theoretically affecting the roughness sensitivityis optimizing the maximum thickness, upper

surface thickness, and transition point

thickness10

(see Figure 3).

Figure 3: Wind turbine airfoil study by Delft University of

Technology

Coefficient of lift is approximated by a polar

analysis of the design airfoil, and is wont to be

maximized in order to produce maximum lift,

but simultaneous tests must be done to take


5/19

into account the total drag on the airfoil, and is

represented by the lift to drag ratio. Many

leading edge roughness design constraints

reduce drag while also reducing lift, so

numerical analysis is often utilized to optimize

all of these parameters.11

3. Assessment Methodology3.1Geographic Study

The GIS data and shapefiles used in this study

are wind speed categories from the nation-wide

from the NREL with direct application to wind

energy availability. The map developed from

this data can give a general idea of which areas

have the most potential for wind power

generation.Further investigation into wind

direction, variability, and other factors like land

management and environmental concerns, is

the next step in location determination. Wind

roses are one source that tracks cardinal

direction as well as speed in those directions.

The wind roses utilized for this study were for

Knoxville, Tennessee. Data was collected

monthly for the year 1961 and 2012 to account

for long term changes in the cyclical pattern. A

sample distributed by hour was taken throughthe months of July and December. This method

allowed a diurnal cycle to be established to the

differences between seasonal patterns to be

investigated.

3.2Airfoil DesignIn designing a unique airfoil for use in this

atypical environment, several factors needed to

be examined. Along with various resources of

study and theory, the program Xfoil developed

by MIT was used as the main design analysis

tool. The design concepts observed were:

delineating an optimum airfoil thickness,

creating an ideal camber line, and maximizing

the lift to drag ratio for peak power generation.

Polars were evaluated for different basis airfoils

of NACA design, with several graphical

parameters changed. The effects of the changes

on the lift to drag ratio and maximum lift were

examined, and the best combination of these

constituents was evaluated.

4. Results and Discussion4.1 Geographic Study

The GIS results from the NREL data can be

found in Figure 4a-c below. From these figures,

it is easy to see that premium wind energy sites

are located along areas of elevation, near the

Cumberland Plateau and Smoky Mountains. For

further elaboration, see the topographical map

of the area in Appendix A. The regions marked

here as upper-division classes (found mostly

along the mountain ridges at the

Tennessee/North Carolina border) have beenidentified as suitable for wind energy

development. According to the NREL wind

energy assessment, the locations found in

mountainous regions are indicative of such

areas as exposed hilltops and ridge crests.

However, as these numbers are based on mean

wind values, they may not indicate high

variability in wind resource in accordance with

local terrain differences. It is probably best to

use this location evaluation as a guideline, and

for exact spots to be scouted and assessedbefore development is undertaken.


6/19

Figure 4a: The GIS data of Tennessee where white

indicates NREL Wind Power Class 1 and black indicatesWind Power Class 7.

Figure 4b: Inset of East Tennessee where white indicatesNREL Wind Power Class 1 and black indicates Wind

Power Class 7.

In addition to average wind speed as

categorized by the NREL study, wind roses for

Knoxville, TN were studied to examine the

temporal variability of the wind resource in this

area. Data from the Western Regional Climate

Center was gathered from the wind roses

available over a period of time. In addition to

annual, seasonal patterns, a diurnal distribution

was also analyzed

In Figure 5, it is easy to see the pattern

of wind behavior over the course of a year.

While average wind speed stays around the

same, the maximum values taper off in the

summer months and gain speed in the winter

months. This is not ideal, as in electricity

demand is typically higher in the summer, and

lower in the winter when households generally

use natural gas or other heating methods to

heat water and spaces. The mean behavior isimportant to note in regards to expectations of

power generation and costs, but the variability

in maximum wind speeds can offer insight into

available power at peak demand.

Figure 4c: Inset of Mid-East Tennessee where white

indicates NREL Wind Power Class 1 and black indicates

Wind Power Class 7.

Figure 5: Average and maximum wind speed data for

Knoxville, TN in 2012.

It is also interesting to note the change

in wind behavior over the course of time in the

long term. The same data available for the year

1961 can be found in figure 6a, and the change

during these years is found in 6b. While the

trends have remained similar, the average

speeds have dropped about 1 m/s across the

board. This may not have much effect on the

feasibility of wind energy at present, but it

should definitely be monitored to examine

whether this trend continues. The effects could

range from eliminating wind energy as a


7/19

possible energy source to other more serious

climatological consequences.

Figure 6a: Average and maximum wind speed data for

Knoxville, TN in 1961.

Figure 6b: Change in average wind speed from 1961 to

2012.

Regarding the daily pattern, Figure 7

shows the hourly variation in two months of the

year 2012. This pattern is more advantageous

for usage in the electricity grid with the highest

wind speeds available in the late afternoon and

evening hours, with the smallest wind speeds

occurring in the dead hours of the morning.

Even across seasons, from the two months

shown (July and December) the patterns

remained similar, suggesting that they were not

significantly affected by seasons. The onenoticeable difference was that the range

throughout the day was lower in summer than

in winter, which is in accordance with the lower

mean wind speed in winter months.

Nevertheless, the diurnal trend present

regardless of seasonal variation matches the

electricity demand. Power need is generally

highest in late afternoon, when children get out

of school and workers come home from their

jobs, and lowest during the day, when no one is

home. These findings can perhaps establish

some argument to the criticism of sporadic

availability in wind as a power source.

Figure 7: Average Daily Wind Speed Variation in July and

December of 2012.

4.2 Airfoil Design

The Xfoil analysis of airfoils designed was

undertaken using the direct analysis method of

the program. There is also an inverse analysis

method, but this study sought to see whatincremental geometrical changes to particular

airfoils did to their aerodynamic performance.

The basis airfoil used in these assessments was

the NACA 6XXXXX series, the airfoil developed

for use in HAWTs, while various other airfoils

were used for comparison purposes.

The first design constraint for analysis

was the maximum thickness, and its effect on

lift, and lift to drag ratio. Five foils (see

Appendix B) based off of the NACA 6XXXXX

series with differing thickness, ranging from21% to 25%, were analyzed for the aerodynamic

performances (Figure 8).


8/19

Figure 8: L/D for six different airfoils ranging in thickness from 0.12 to 0.25

As is obvious from Figure 8, the

maximum thickness of the airfoil significantly

affects the lift to drag ratio. This relationship

was outlined before, in stating that various

design components would be beneficial to

decreasing sensitivity to roughness, but

detrimental to both maximum lift and lift to

drag ratio. It is, however, possible to optimize

the effect, while keeping both positive aspects.It is perhaps hard to tell in these figures, but for

the smaller thicknesses, the L/D relationships

are very close, almost overlapping. As the

thicknesses increase, the changes become more

drastic, indicating that there is a point where

the benefit gained in insensitivity to roughness

through increasing thickness is surpassed by the

negative effects in the lift to drag ratio. It is

more apparent, as seen in Figure 9 and 10 that

the maximum lift and maximum lift to drag ratio

are adversely affected by the increase inthickness. While L/D remains fairly constant

until a thickness value of about 0.18, where it

drops off steeply, the lift behavior is more

extreme, but consistent with the previous

findings.

Figure 9: Maximum lift to drag ratio of the six airfoils of

varying thickness found in Appendix B.


9/19

Figure 10: Maximum lift coefficient of the six airfoils of

varying thickness found in Appendix B.

In this series of airfoils, this point of contention

was found to be around 18%, although otherstudies have found the best thickness to be

about 21%.

The next design constraint analyzed was

the behavior of the camber line, or more

specifically, the high point of the camber line. A

series of airfoils was developed with increasing

camber positions, a 610XX, 620XX, 630XX, and

640XX series were all compared with respect to

aerodynamic properties at stall. In improving

sensitivity to surface roughness, the goal is to

make the laminar to turbulent behavior of the

airflow as close to the leading edge as possible,

at angles of attack post-CLmax. In Appendix C, the

flow behavior across the airfoils at an of 15 is

shown. At a higher asymmetry across the airfoil,

it is shown that the transition of laminar to

turbulent flow is moved closer to the leading

edge. Therefore, increasing the camber line can

be said to increase insensitivity to roughness.

However, this design point is again contrary to

increasing L/D, so an optimization of both

components would be necessary. A numerical

analysis is ideal, but a general value of 10%

camber has been accepted (NACA 62021).

Finally a design implementation isnecessary to maximize lift available. An angling

of the trailing edge of the airfoil (as seen in

Appendix D) was found to significantly increase

the lift across angles of attack, when all other

things were equal, see Figure 9.

Figure 11: This chart shows the increase in lift as a result of a trailing edge modification on the NACA 62021 series airfoil.


10/19

From the pressure vector graphs found in

Appendix C, it can be inferred that the trailing

edge modification significantly increased the

magnitude of the vector across the top of the

airfoil, and mostly decreased CP across the

bottom, leading to an increase in maximum lift.

The increase under the new trailing edge did

not apparently affect the lift generated,

especially at low angles of attack.

The resultant airfoil takes into account

the previous conclusions in its design, seen in

Figure 12. Based off of the NACA 62018, the

optimum camber and thickness are called

inherently in NACA form, but the tail rotation is

graphically customized.

Figure 12: Graph of tailored airfoil

The airfoils polar assessment can be found in

Figure 13, resulting in a maximum lift coefficient

at about 1.74, at an angle of attack of around10.2. This is higher than any of the individual

test cases.

Figure 13: Aerodynamic Performance

Figure 14: CPV of tailored airfoil.

Figure 14 shows a pressure vector display of the

airfoil at maximum lift ( = 10.2). There is

clearly an indication of positive lift, the pressure

vectors along the top of the airfoil, while drag,

the components along the bottom, has been

relatively minimized.


11/19

4.3 Environmental Assessment

While wind energy developments may be less

destructive than fossil fuels in terms of

emissions and other spectrums of

environmental friendliness, but they are not

without fault. Factors that influence the

intrusiveness of wind farms include noise

pollution, effects on wildlife and habitat, as well

as impacts on land and seascapes in regards to

tourism and community value.

Though studies such as those outlined

in Table 2 have shown that bird, bat, and other

avian fatalities are next to negligible and more

obvious in the short term compared to other

human and natural factors,12

it is important to

keep these impacts in mind and to minimize

them as much as possible. This can be managedthrough conscious siting practices, including

thorough habitat and migratory avoidance, and

operation procedures like ceasing processes in

bad weather. For Eastern Tennessee on

exposed, rocky ridges in particular the focus of

this study, special attention must be paid to

migratory bat population. During migratory

season (late summer) bat mortality rate has

been shown to climb as high as 20 bats per

turbine per year13

at the currently active Buffalo

Mountain Wind Farm. If migratory patternsremain similar, however, energy production

could be planned around bat population

movement, as summer nights have been shown

above to be some of the least wind power

dense periods throughout the year. Turbine

type is also a significant aspect of aviary

mortality. For example, lattice type tower

structures provide roosting prospects for local

wildlife, attracting rather than discouraging

birdlife from interaction with the wind turbines.

Further, VAWTs have been shown to be safer

than prop-type turbines in bird mortality rates,

for their more easily avoidable structural

pattern.

Table 2: From Saidur, et al, this table outlines the relative

danger that wind turbines pose to wildlife as compared

to other human-related causes.

The next most common concern with respect to

wind is noise pollution, possibly resulting in

decreased property values and creating

unpleasant environments for residential areas.

An obvious solution to this problem is to site

wind farms at a clear distance from areas that

might be affected. In this studys focus area,

this can be achieved easily, as most regions ofEast Tennessee and the surrounding

Appalachian range are sparsely populated.

While higher infrastructure prices would be one

negative consequence of this solution, this is

already normal with most energy generating

methods. Coal, natural gas, and crude oil are all

mined and refined at considerable distances

from their demand locations. A second, and

perhaps more fitting option to consider, is the

utilization of VAWTs, like the Darrieus and H-

Rotor type of turbine. These create less noise,

14

and if chosen at a small size, would be less

environmentally impactful to wildlife. VAWTs

are not typically implemented because of their

low density in power generation, but because of

the smaller wind speed availability in this area

would perhaps be a good tradeoff in terms of

environmental impact.

5. ConclusionIn covering a wind feasibility study in this region

taking into account geographic and

meteorological patterns, unique airfoil design,

and environmental assessment, this study has

been slightly exhaustive on certain points.

Nevertheless, certain topics could be furthered.

The goals elaborated in the airfoil design would

best be suited to a numerical analysis, with

weights placed on factors most important to


12/19


13/19

6. Appendix A


14/19


15/19


16/19

8. Appendix CNACA 61021

NACA 62021

NACA 63021

NACA 64021


17/19

9. Appendix DNACA 62021

NACA 62021B


18/19

NACA 62021C

NACA 62021D


19/19

1Rehman, Shafiqur. (2003). Wind energy resources assessment for Yanbo, Saudi Arabia. Energy Conversion &

Management, 45, 2019-2032.

2

"Wind Energy Resource Atlas of the United States" Renewable Resource Maps and Data. National RenewableEnergy Lab. .

3Mohammadi, Kasra; Mostafaeipour, Ali. (2012). Using different methods for comprehensive study of wind turbine

utilization in Zarrineh, Iran. Energy Conversion & Management, 65, 463-470.

4Eriksson, Sandra; Bernhoff, Hans; Leijon, Mats. (2006). Evaluation of different turbine concepts for wind power.

Renewable & Sustainable Energy Reviews, 12, 1419-1434.

5"Wind Energy Resource Atlas of the United States" Renewable Resource Maps and Data. National Renewable

Energy Lab. .

6 Ramachandra, T.V.; Shruthi, B.V. (2004). Wind energy potential mapping in Karnataka, India, using GIS. Energy

Conversion & Management, 46, 1561-1578.

7Baban, Serwan M.J.; Parry, Tim. (2001). Developing and applying a GIS-assisted approach to locating wind farms

in the UK. Renewable Energy, 24(1), 59-71.

8Dahl, Kristian S.; Fuglsang, Peter. Design of the Wind Turbine Airfoil Family RISO-A-XX. Riso National Laboratory.

9"Wind Turbines" CATs: Coherent Application Threads. Boston University Mechanical Engineering.

10Somers, D.M. (1998). Effects of Airfoil Thickness and Maximum Lift Coefficient on Roughness Sensitivity.

11Dahl, Kristian S.; Fuglsang, Peter. Design of the Wind Turbine Airfoil Family RISO-A-XX. Riso National Laboratory.

12Saidur, R. (2011). Environmental impact of wind energy. Renewable & Sustainable Energy Reviews, 15, 2423-

2430.

13Fiedler, Jenny K. (2004). Assessment of Bat Mortality and Activity at Buffalo Mountain Windfarm, Eastern

Tennessee.

14Eriksson, Sandra; Bernhoff, Hans; Leijon, Mats. (2006). Evaluation of different turbine concepts for wind power.

Renewable & Sustainable Energy Reviews, 12, 1419-1434.

Wind Feasibility and Optimization Study in East Tennessee

Documents

Transcript of Wind Feasibility and Optimization Study in East Tennessee