Design and implementation of wind energy system in Saudi Arabia

Renewable Energy 60 (2013) 42e52

Contents lists available at SciVerse ScienceDirect

Renewable Energy journal homepage: www.elsevier.com/locate/renene

Technical note

Design and implementation of wind energy system in Saudi Arabia Ali M. Eltamaly Sustainable Energy Technologies Center, Electrical Engineering Dept., King Saud University, P.O. Box 800, Riyadh 11421, Saudi Arabia

a r t i c l e i n f o

a b s t r a c t

Article history: Received 9 November 2012 Accepted 7 April 2013 Available online 11 May 2013

This paper introduces an accurate procedure to choose the best site from many sites and suitable wind turbines for these sites depending on the minimum price of kWh generated (Energy Cost Figure (ECF)) from wind energy system. In this paper a new proposed computer program has been introduced to perform all the calculations and optimization required to accurately design the wind energy system and matching between sites and wind turbines. Some of cost calculations of energy methods have been introduced and compared to choose the most suitable method. The data for five sites in Saudi Arabia and hundred wind turbines have been used to choose the best site and the optimum wind turbine for each site. These sites are Yanbo, Dhahran, Dhulom, Riyadh, and Qaisumah. One hundred wind turbines have been used to choose the best one for each site. This program is built in a generic form which allows it to be used with unlimited number of sites and wind turbines in all over the world. The program is written by using Visual Fortran and it is verified with simple calculation in Excel. The paper showed that the best site is Dhahran and the suitable wind turbine for this site is KMW-ERNO with 5.85 Cents/kWh. The worst site to install wind energy system is Riyadh with minimum price of kWh of 12.81 Cents/kWh in case of using GE Energy 2 wind turbine. Ó 2013 Elsevier Ltd. All rights reserved.

Keywords: Wind energy system design Matching between site and wind turbine Cost of energy Price minimization of the generated kWh

1. Introduction Wind energy applications require open area or available shores for wind energy plants. Saudi Arabia is a vast country with wide open areas and long shores. The wind speed in most of these areas is high enough to make the application of wind energy economical. Saudi Arabia authorities recognize the importance of renewable energy, especially, wind, and they will invest billions in this promising sector of power. Even though Saudi Arabia has huge resources of oil, it is keenly interested in taking an active part in the development of new technologies for exploiting and utilizing renewable sources of energy [1]. The electricity production from wind will save oil that can be exported for increasing national income. Also, the production of electric power from wind energy will reduce environment pollution that could be generated from conventional power plants. Recently, a lot of researches in the evaluating the applications of wind energy systems in Saudi Arabia are introduced. Most of these researches recommend wind as a promising and economical source of energy in Saudi Arabia [2e10]. While the wind resource potential in Saudi Arabia is significant, there are many issues surrounding its development. These include the intermittency of the resource, its seasonal and diurnal

E-mail addresses: [email protected], [email protected]. 0960-1481/$ e see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.renene.2013.04.006

characteristics, its geographically remote locations, and the electrical grid infrastructure that must be used to transmit the wind energy to load areas. All of these issues pose significant technical barriers to the full development of Saudi Arabia’s wind potential. The matching between the site and wind turbine has been introduced in many literature [11,12]. Most of these researches did not take into account many important issues like the suitable economical situation in Saudi Arabia and the energy balance between the required load and the generated power. Also, the market available software such as Homer, Retscreen, and etc are not flexible enough to change the cost calculations and it is not able to extract much information as those available from the new proposed computer program. This paper introduces an accurate procedure to choose the best site from many sites and suitable wind turbines for these sites depending on the minimum price of kWh generated from wind energy system. In this paper a new proposed computer program has been introduced to perform all the calculations and optimization required to accurately design the wind energy system and matching between sites and wind turbines. Starting any program of using wind energy system in a utility scale in any country in the world requires many steps. The first step is to collect accurate wind speed data for many sites to be used in the selection process of the best site. The second step is to determine the load curves which required to be supplied. The third step

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52

is to collect the data of available wind turbines in the market. This leads to the question ‘What is the most appropriate wind turbine for a particular site?’. The aim of this research is to provide an accurate answer to this question. A new proposed computer program is designed in this paper to answer this question and many other questions. The function of the proposed program is to decide which the best site from many available sites is, and which the suitable wind turbine for each site and the minimum price for generated kWh. This decision requires very accurate statistical calculations. Another information can be extracted from this computer program such as the Weibull parameters, capacity factor and other information that may be helpful for researchers. The steps of these processes are very long and require long time and great efforts for calculations and optimizations. The new proposed computer program will solve this problem in a very short time and gives accurate results in a flexible fashion. This computer program can be applied in any region in the world because it has the flexibility to change the economic calculation to suit any country and it is able to handle unlimited numbers of sites and wind turbines. 2. Design of the proposed computer program The flowchart of the computer program is shown in Fig. 1. The program has a main part and five subroutines. Each subroutine will perform a certain function as shown in the following sections. This program has been applied to the five sites in Saudi Arabia. The performance data of hundred of market available wind turbines were used to select the most suitable one for each site. 3. Program input data Wind speed variation of the site and the performance characteristics of wind turbines are the main factors that affect the performance of wind energy system and affect the cost of kWh generated from it. The data required for the program are:

1. Hourly wind speed data for Yanbo, Dhahran, Dhulom, Riyadh, and Qaisumahsites of Saudi Arabia will be used. The hourly wind speed data of these sites will be processed using statistical procedures. The computer program can perform the optimization for unlimited number of the available data sites. Wind data is collected from many sources as metrological authority and over the internet. Wind speeds can be collected for many sites for different period of times. Meteorologists generally conclude that it takes at least 5 years of wind data to determine a reliable average and variance of the wind speed. Some researchers claim that shorter period of time may be acceptable for designing renewable energy system with acceptable confidence [12]. It is better to have a small interval between each reading of the wind speed data. Thirty minutes are recommended interval between each two points of data. But, this may not available for all sites under study because some of these sites have one-hour interval. So, the interval used in this research paper is one-hour to fit all available data. 2. Performance data for market available wind turbines are introduced, such as rated power, hub height, diameter of swept area, cut-in speed, rated speed, cut-out speed, price of wind turbine, and efficiency of the mechanical and electrical system. In this step hundred market available wind turbines are introduced to the computer program. The computer program can perform the optimization for unlimited number of wind turbines. 3. Hourly loads required to be supplied from the wind energy system. The load data used in this program is actual data for small city in Saudi Arabia. The average power required for this load is 22.5 MW.

4. Steps of the proposed computer program The main computer program reads all the data mentioned in previous section. After that, the flow will go to subroutines to make the calculations, comparisons, and optimizations. Each subroutine makes a certain function and then sends its results to the next subroutines. The function of each subroutine is illustrated in the following:

Reading the data Wind turbine parameters Hourly wind speed Hourly load power

4.1. Weibull parameters calculations

Weibull Parameters Calculation, (c and k)

The purpose of this subroutine is to determine the Weibull, scale and shape parameters, c and k. A good estimation for c and k can be obtained quickly as the following [13]:

Calculating CF, Pav, and ANWTG

c ¼ 1:12U

Energy Balance Subroutine

k ¼ yes

Use the data for new site

Is there any other WTGs? yes

ð1:5  k  3:0Þ

(1)

Also, if the mean and variance of the wind speed are known, then approximation for k from Ref. [13] can be used as shown in (2);

Cost Estimation

Use the data for new WTG

43

no


s 1:086

(2)

U

The variance of the Weibull density function can be shown to be:

Is there any other sites? no Output Results

Fig. 1. Simple block diagram of the computer program.

 

s2 ¼ c2 G 1 þ

# "    Gð1 þ 2=kÞ 2 1 1  G2 1 þ ¼ ðUÞ2 2 k k G ð1 þ 1=kÞ (3)

This is a reasonably good approximation over the range 1  k  10. Once k has been determined, c can be obtained as the following:

44

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52

c ¼

U

(4)

Gð1 þ 1=kÞ

A relation between c and k and other parameters such as average wind speed and hub height is extracted from the results of this subroutine. Another method depends on the accurate statistical analysis for obtaining Weibull parameters has been used in this computer program [14]. The final results for the Weibull parameters are:

k ¼ a c ¼ expðb=kÞ

(5)

ho ¼ Cp hm hg

(13)

If Pe > PL ; Then PT ¼ Pe  PL and Pf ¼ 0

(14)

If Pe > PL ; Then PF ¼ PL  Pe and PT ¼ 0

(15)

For energy balance the following conditions must be satisfied: 8760 X i¼1

Pe ðiÞ 

8760 X

PL ðiÞ ¼ 0;

i¼1

8760 X

PT ðiÞ ¼

i¼1

8760 X

PF ðiÞ ¼ 0

(16)

i¼1

where

Pw

Pw a ¼

i ¼ 1 xi yi 

i ¼ 1 xi

Pw

i¼1

w Pw

!2

yi

Pw ¼

i ¼ 1 ðxi

i ¼ 1 xi

Pw

2 i ¼ 1 xi



4.4. Energy price calculation

Pw

 xÞ i ¼ 1 ðyi  yÞ Pw 2 i ¼ 1 ðxi  xÞ

This subroutine is used to compute the ECF for each site and each type of wind turbine. 4.5. ECF optimization subroutine

w (6)

w w 1 X a X y  x b ¼ yi  axi ¼ w i¼1 i w i¼1 i

(7)

5. Economic analysis method

and

yi ¼ lnð  lnð1  Fðui ÞÞÞ; xi ¼ lnðui Þ

(8)

4.2. Capacity factor and average number of wind turbines calculation The purpose of this subroutine is to determine the Capacity Factor, CF, and the average number of wind turbine generators, ANWTG. The capacity factor can be obtained from the following equation [13]:

CF ¼

i h i h exp  ðuc =cÞk  exp  ður =cÞk k

ður =cÞ  ðuc =cÞ

k

h
k i  exp  uf =c

The average power of wind turbine can be calculated as:

Peave ¼ CF $Per

(10)

The average number of wind turbine generator can be calculated as:

ANWT ¼

PLav Peave

(11)

There are two simplified methods are discussed in the following sections. These methods provide fast feasibility study for installing wind energy system and it can be used to make matching between site and wind turbines depending on the lowest price of the generated kWh. These methods are representing simple but not accurate techniques to calculate the price of the kWh generated from wind energy system. 1) Simple payback period analysis A payback calculation compares revenue with costs and determines the length of time required to recoup an initial investment. The simplest payback period (in years) can be obtained from the following equations [15]:

4.3. Energy balance subroutine The purpose of this subroutine is to determine the optimal number of wind turbines required and the yearly energy generated from the wind energy system. Then the optimum number of wind turbines required and the energy output for each case can be obtained. The output power from wind energy system is given by:

1 r*At *u3 *Nt *h0 2

The economic analysis proposed in this section is used to estimate the price of the generated unit energy from wind energy system. The estimation of the price of the generated kWh depends on the accurate estimation of the wind speed data and the costs of the components and the operating and maintenance costs. The general purpose of such methods is not only to determine the economic performance of a given design of wind energy system, but also to compare it with conventional and other renewable energy based systems and to match between the site and wind turbine depending on the minimum energy price. The following are different methods of overall economic analysis: 5.1. Simplified economic analysis methods

(9)

Pe ¼

The purpose of this subroutine is to select the minimum value of ECF and then determine the corresponding site and wind turbine type.

(12)

SP ¼ CC =AAR

(17)

AAR ¼ Ea PE

(18)

SP ¼ CC =ðEa Pe Þ

(19)

It should be pointed out that the calculation of simple payback period omits many factors that may have a significant effect on the

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52

system economic cost effectiveness. These include escalating fuel (in a hybrid power system) and loan costs, depreciation on capital costs, operation & maintenance costs (O&M), and variations in the value of delivered electricity. Some of these variables are attempted to be included in some author’s calculations for a simple payback period [16e19]. This method is a simplest method and it takes short time to do the calculations and get the preliminary results and information. The analysis team uses SP of 10 years or less to avoid assuming values for energy escalation rates and O&M inflation factors that are required for other life-cycle cost values [20]. 2) Cost of energy analysis The cost of energy (COE) is defined as the cost of the kWh generated from wind energy system. That is:

Cost of energy ¼

Operating costs Energy produced

PV ¼

45

N X A A A 1 þ :::::: þ ¼ A þ N j 1 þ r ð1 þ rÞ2 ð1 þ rÞ j ¼ 1 ð1 þ rÞ

(25)

3) Capital recovery factor The capital recovery factor (CRF) is used to determine the amount of each future payment required to accumulate a given present value when the discount rate and the number of payments are known. Using Equation (25), the capital recovery factor is given by Ref. [15]:

 CRF ¼

h i r= 1  ð1 þ rÞN ; if rs0

(26)

1=N if r ¼ 0

(20) 4) Net present value

The simplest calculation of COE is given by Ref. [15]:

COE ¼ ½ðCC *FCRÞ þ CO&M =Ea

(21)

The fixed charge rate, FCR will generally reflect the interest on pays or the value of interest received if money were displaced from savings.

PV ¼ C=ð1 þ rÞj

Life-cycle costing (LCC) is a commonly used method for the economic evaluation of energy producing systems based on the principles of the ‘time value’ of money. The following parameters are included in the LCC analysis: 1) Time value of money and present worth factor

NPV ¼

N X

C

j ¼ 1 ð1

þ rÞj

(28)

If the cost C is inflated at an annual rate i, the cost Cj in year j becomes [15]:

(29)

Thus, the net present value, NPV, becomes:

NPV ¼

 N  X 1þi j j¼1

1þr

C

(30)

(22)

The savings version of net present value, NPVs is defined as follows [15]:

(23)

NPVs ¼

 N  X 1þi j j¼1

1þr

ðS  CÞ

(31)

If only cost factors are considered, then a cost version of net present value, NPVC, may be used. NPVC may be found from the following equation [15]:

2) Levelizing Levelizing is a method for expressing costs or revenues that occur once or in irregular intervals as equivalent equal payments at regular intervals. Considering a loan of value PVN is to be repaid with a single payment FN at the end of N years. The payment is [21]:

FN ¼ PVN $ð1 þ rÞN

PVi ¼

j¼1

The present worth factor, PWF is given by Ref. [15]:

PWF ¼ PV=FV ¼ ð1 þ rÞ

N X

Cj ¼ Cð1 þ iÞj

A unit of currency that is to be paid (or spent) in the future will not have the same value as one available today. This is true even if there is no inflation, since a unit of currency can be invested and bare interest. Thus its value is increased by the interest. The future value, FV, after N years is [15]:

N

(27)

Thus, the NPV of a cost C to be paid each year for N years is [15]:

5.2. Life-cycle costing methods

FV ¼ PVð1 þ rÞN

The net present value (NPV) is defined as the sum of all relevant present values. From Equation (22), the present value of a future cost, C, evaluated at year j is [15]:

(24)

A loan that is to be repaid in N equal installments can be considered as the sum of N loans, one for each year, the jth loan being repaid in a single installment A at the end of the jth year. Thus, the value, PV, of the loan equals [21]:

 NPVC ¼ pd þ pa Y

1 ;N 1þr



 þ CC fOM Y

1þi ;L 1þr

 (32)

Levelized cost of energy (COEL) for the utility-based calculation of cost of energy can be given by Ref. [15]:

P COEL ¼

ðLevelized annual costsÞ Annual energy production

(33)

Using CRF and NPVC the COEL can be calculated by Ref. [5]:

COEL ¼



ðNPVC CRF Annual energy production

(34)

46

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52

5.3. Electric utility based economic analysis

I ¼ In the United States, electric utilities and the wind industry commonly use either of following two methods to estimate the COE from a utility-sized wind energy system: 1) EPRI TAG method This method produces a simplest form for wind energy systems; COE is calculated as [22]:

0

 Cc

1

 B C COE ¼ FCR @ A þ CO&M 8760,CF

(35)

Since this method produces a levelized energy cost it can be applied to a number of technologies, including conventional power plants (with the addition of fuel costs) for a useful comparison index. Some limitations of the EPRI TAG method include that it assumes a debt term life equal to the life of the power plant and it does not readily allow for variable equity return, variable debt repayment, or variable costs.

j X

The cash flow method is based on the use of an accounting type spreadsheet that requires an annual input of estimated income and expenses over the lifetime of the project [23,24]. The cash flow method allows for the real variations that can be expected in cost, operational, and economic data, such as price increases, inflation, and changing interest rates.

(38)

i¼1

O&M costs depend on the number of wind turbines, the wind turbine type and the site conditions. This method recommends project specific estimates of the O&M costs to be specified for each year of the scheme’s lifetime. Social costs (SC) may be associated with environmental damage, nuisance to people, etc. Sometimes social costs of wind energy production are small or negligible, especially when compared to those associated with energy generation from non-renewable sources. Replacement cost (RC) or major repairs during the adopted lifetime should be evaluated, and it is recommended that project specific estimates are made of the timing and cost of possible major repairs. The salvage value (SV) is defined as the difference between the scrap value and the decommissioning cost of the entire scheme at the end of the lifetime adopted for the economic analysis. The discount rate (r) defined as the rate at which the nominal rate, i, exceeds the inflation rate, v, i.e. [25]:

1þr ¼

2) Cash flow method

Ii $ð1 þ rÞti

1þi 1þv

(39)

The relation between the annual utilized energy (AUEt) and the annual net energy (ANEt) is described as [26]:

AUEt ¼ ANEt $Klost; t $Kutil; t AUEt ¼ ANEt $Klost; t $Kutil; t ¼ Epot $Kper; t $Ksite; t $Kava; t $Klost; t $Kutil; t

5.4. Levelized production cost (LPC) In this method the cost components are assumed to be the investment cost, operation and maintenance cost, repair cost, salvage value and social cost. The following sections introduce two approaches to calculate the levelized production cost: 1) First approach In this approach it is assumed that all costs are given in a fixed currency for a specified year. The currency and cost level year should be decided and clearly declared by the assessor when reporting the estimated cost of energy. In these calculations all costs are discounted to the present value. The discounted present value of the total cost (TC) is given as [25]:

TC ¼ I þ

n X

ðO&Mt þ SCt þ RCt Þ$ð1 þ rÞt  SV$ð1 þ rÞn

ZN

Epot ¼ 8760$ pðuÞ$f u du

LPC ¼ TC=

n X

AUEt $ð1 þ rÞt

(37)

t¼1

In many cases one or more of the input parameters of this approach will be known explicitly, and of course, the known figures should be used whenever possible. The investment (I) should include all the costs of constructing the wind energy conversion systems. The total investment can be calculated by Ref. [25]:

(42)

0

The wind speed distribution f(u) should ideally be based on many years of on-site wind speed measurements, but in practice it will often be necessary to extrapolate long term wind data from nearby high quality measurement stations, using for instance the wind atlas method [27], or by applying the statistical “measurecorrelate-predict” approach [28]. The power curve p(u) normally gives the net power output for standard air density conditions and for carefully selected weather. For a stall regulated wind turbine, p(u) can be calculated by Ref. [29]:

(36) The levelized production cost is given as [25]:

(41)

The annual potential energy output (Epot) of a wind turbine experiencing specific meteorological conditions is given by Ref. [25]:

p u ¼ pðuÞstd $

t¼1

(40)

r 1:225

(43)

The performance of a wind turbine may be reduced due to dirt, rain or ice on the blades. Cleaning of the blades must be included in the O&M costs or a reduction in the annual energy output DEper, t relative to the potential output must be assumed. The wind turbine performance factor Kpert, t can be calculated as [30]:

Kpert; t ¼ 1 

DEper; t Epot

(44)

In some cases the site surroundings may change with time due to erection of new wind turbines, tree planting, construction of new houses, etc. thus influencing the wind speed distribution and the

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52 2%

2% 2%

2% 2%

5%

DEsite; t Epot $Kper; t

7%

Ksite; t ¼ 1 

8%

energy output from the wind turbine. In such cases, the reduction in annual energy output, DEsite, t,due to the changed surroundings should take into account. The annual reduction may be expressed by means of the site factor Ksite, t as [30]:

47

(45)

70%

The technical availability factor Kava, t is defined by the energy loss DEava, t due to the wind turbine availability as:

Kava;t ¼ 1 

DEava;t

(46)

Epot $Kper;t $Ksite;t

The annual electrical transmission losses factor Klos, calculated as [25]:

Klos; t ¼ 1 

t

can be

DElos; t

(47)

ANEt

Kutil is the utilization factor and defined as [25]:

Kutil; t ¼ 1 

DEutil; t ANEt $Klos; t

(48)

2) Second approach This approach assumes that the annual utilized energy to be constant from year to year (i.e. AUEt ¼ AUE for t ¼ 1 to n). In such case, the LPC can be calculated as [25]:

LPC ¼ I=ða$AUEÞ þ TOM=AUE

(49)

a is the annuity factor and is defined as [25]:

a ¼ 1=

n X

ð1 þ rÞt ¼




. r 1  ð1 þ rÞn

(50)

t¼1

TOM is the total levelized annual “down line costs” and is defined as [25]:

TOM ¼ a1 $

n X

ðOMt þ SCt þ RCt Þð1 þ rÞt  SVð1 þ rÞn

t ¼1

(51)

Wind Turbines

Electrical Installations

Grid Connections

Civil Work

Land Rent

Project Management

Insurance

Consultancy

Financial Cost

Fig. 2. Fair range of the cost share of different components of the wind energy system for commercial size wind turbine.

wind energy system components, but an approximate cost analysis methodology can be used in this paper. This methodology was applied to determine the cost per kWh in each type of wind turbine and each site. This method is simple, efficiently and used to compare the cost of energy from alternative generating devices. From the fourth subroutine the total price of kW generated by wind turbine approximately equal to $700 per kW (based on year of 2010). The total price of microprocessor is $2.3 per kW, the total price of main substation is $10.4 per kW, the total price of modem for remote control in central control station is about $4.16 per kW and the total price of transmission line is about $1.3 per kW. Then:

TPWTG ¼ $700*NWTG*Pr

(52)

TPMIC ¼ $2:3*NWTG*Pr

(53)

TPMS ¼ $10:4*NWTG*Pr

(54)

TPCCS ¼ $4:16*NWTG*Pr

(55)

TPTL ¼ $1:3*NWTG*Pr

(56)

The cost of the operation and maintenance is about 10% of the total cost. Then:

Total Price ¼ 1:1*ðTPWTG þ TPMIC þ TPMS þ TPRC þ TPCCS þ TPTLÞ (57)

6. Economic performance sensitivity analysis The previous sections have described a number of techniques to determine the economic performance parameters that can be used to evaluate various wind systems, or to compare their performance with other types of power system. These economic techniques have different evaluation ideas and different assumptions. One of these techniques may be suitable for certain project and the other may not suitable. The cost share of different components of the wind energy system for commercial size wind turbines (from 2 to 5 MW) is shown in Fig. 2 [31,32]. This price range may differ from country to country. The cost of kWh produced is very important for the consumers and power companies to return a profit on the capital invested. It is very difficult to determine an accurate cost of the kWh generated due to the variation of wind speed and changing in the price of

Total Price ¼ 1:1*ð718:16*NWTG*PrÞ ECF ¼

Total Price*LF YE*0:9

(58)

(59)

At 12% interest and 10 year recovery time, LF ¼ 0.177.

7. Summary of the output results Many results have been extracted from the proposed computer program. Each part in the computer program can provide the researchers with a lot of helpful information due to the wide input. The results of this computer program have been checked with the results obtained from software packages used in this application such as Homer and Retscreen. The following is the summary of the output results:

48

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52

16

Shape Parameter, k

15

Rated wind speed, ur

3.5

data 2 linear

14 13 12

Yanbou Yanbou Dohloum Dohloum Dahhran Dahhran Riyadh Riyadh Qaysoma Qaysoma

3

2.5

11 10

2

2.5

3

3.5

2 40

4

60

80

100

120

Cut-in wind speed

180

200

220

240

Fig. 5. A relation between the shape parameter, k along with the hub height, h of wind turbines for five sites under study.

Fig. 3. The relation between the rated and cut-in wind speeds.

1) From the data of one hundred wind turbines, the relation between the rated and cut-in wind speed is shown in Fig. 3. The relation between the rated wind speed and cut-in wind speed is obtained by linear interpolation and the equation is shown in (60). This curve shows design constraints for the rated and cut in wind speed.

Vr ¼ 1:2679 Vc þ 8:5227

140 160 Hub Height

(60)

It is clear from Fig. 3 that the rated wind speed is directly proportional to cut-in wind speed. Also, the minimum cut-in and rated wind speed is 2.5 m/s and 10 m/s respectively. The maximum cut-in and rated wind speed is 4 m/s and 15 m/s respectively. Relations between scale and shape parameters along with the hub height, h of wind turbine are shown in Fig. 4 and Fig. 5 respectively. Fig. 4 shows that the value of scale parameter, c is directly proportional to the hub height, h. This relation is very useful in case of it is required to obtain the value of scale parameter, c at any height. It is also clear from Fig. 5 that the shape parameter, k is almost constant for different hub heights. Also it is clear that the value of shape parameter, k is directly proportional to the hub height with very low slope where the change in value of k is less than 5% in the range of hub height. The relation between the scale parameter and average wind speed for three-sites under study is shown in the following Fig. 6. It is clear from this figure that the scale parameter, c is directly proportional to the average wind speed of the site, V for all sites. Also, it is clear that the linear relation between scale and average wind speed lie on a linear relation in ideal way for all sites. So it is easy to

get the scale parameter for any site from its average wind speed from the following relation:

c ¼ 1:1064*Uav  0:49812

(61)

Many references have introduces a relation between scale parameters and average wind speed. O.A. Jaramillo [33] uses Gamma function to predict this relation as shown in (62). The results obtained from Ref. [33] are typically aligned on the curve shown in Fig. 11. G. L. Johnson [34],E. L. Skidmore [35] gives another linear relation between scale parameters and average wind speed as shown in (63).

  Uav 1 ¼ G 1þ c k

(62)

c ¼ 1:12*Uav

(63)

The relation between the shape parameter and average wind speed for five-sites under study is shown in Fig. 7. It is clear from this figure that the shape parameter, k is directly proportional to the average wind speed of the site, Uav but the change in shape parameter; k is very limited with change in average wind speed. The relation between shape and scale parameters for five sites and one hundred wind turbines under study is shown in Fig. 8. From this figure it is clear that, the relation is linear but differs from site to site. This relation is shown in many references [35] as a single

10 C=1.1064*uav-0.49812 9

8.5

Scale Parameter, C

8

Scale Parame te r,C

7.5 7

Yanbou Yanbou

6.5

Dohloum Dohloum Dahhran Dahhran

6 5.5

Riyadh Riyadh

5

8 7 6 5

Qaysoma Qaysoma

4.5 40

60

80

100

120

140

160

180

200

220

240

Hub Height Fig. 4. A relation between the scale parameter, c along with the hub height, h of wind turbines for five sites under study.

4 4

5

6 7 8 Average wind speed

9

10

Fig. 6. The relation between scale parameter and average wind speed for five-sites and one hundred wind turbines under study.

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52

49

Shape Parameter, k

3.5 Yanbo Yanbo Douhlom Douhlom Dahhran Dahhran Riyadh Riyadh Qayssuma Qayssuma

3

2.5

2 4.5

5

5.5

6

6.5 7 7.5 Average wind Speed

8

8.5

9

Fig. 7. The relation between shape parameter, k and the average wind speed for five sites and one hundred wind turbines.

relation as shown in the following equation (64) [35] which is not correct for all sites.

k ¼ 0:52 þ 0:23c

(64)

The generation of electrical energy by wind turbine at a specific site depends upon many factors. For a given wind site, the following parameters should be known or can be calculated, the mean wind speed and Weibull scale parameter (c) and shape parameter (k). Also, for specific wind turbine the performance parameters should be known that include: cut-in (uc), rated (ur), and furling (uf) wind speeds, the hub height, and the rated power. The rated speed is the most important for wind turbine deign. If the rated speed is chosen too low to utilize the low wind speeds, much energy will be lost in the higher wind speed. On the contrary, if the rated speed is too high, the turbine seldom operates at its rated capacity and also will lose too much energy at lower wind speed. This means the rated speed has to be selected such that the turbine yields higher energy. To achieve this condition a suitable relation between the site parameter (such as average wind speed) and rated wind speed (as wind turbine parameter) with the capacity factor, CF is obtained to get the condition for maximum capacity factor as shown in Fig. 9. This relation is drawn for five sites and one hundred wind turbine under study. From this figure it is clear that the capacity factor is increasing considerably with increasing the average wind speed and slightly increasing with reducing the rated speed. This is correct in the normal operating range. In extending the operation of the wind turbine beyond the normal operating range, the capacity factor increases with increasing the ratio of average and rated wind speed till Uav/ur

Fig. 9. The relation between average wind speed of site, rated wind speed of wind turbine and the capacity factor, CF.

equal approximately to 1.22 as shown in Fig. 10. After this value any increase in the ratio Uav/ur will reduce the capacity factor of the wind turbine. The critical value of 1.22 is the average condition for the five sites under study. The normal operating is always below this value in most cases. The results of energy price along with the capacity factor of the five sites are shown in Fig. 11. It is clear from this figure that the energy price inversely proportional to the capacity factor. The energy unit price in Cents/kWh for five sites along with average wind speed and Uav/ur and one hundred wind turbines is shown in Fig. 12 and Fig. 13 respectively. It is clear from this figure that the price of generated kWh is inversely proportional to Uav/ur. The price of kWh generated from five sites and the best five out of one hundred wind turbines are shown in Table 1 and Fig. 14. The best wind turbine for each site and the price of kWh generated in this matching is shown in Table 2. Its clear from these tables that the lowest price for kWh is stand for Dhahran site and KMW-ERNO wind turbine which is 5.85 (Cents/kWh). So, it can be said that the best site for wind energy system installation is Dhahran and the best wind turbine is KMW-ERNO. This wind turbine also is the best wind turbine for Qaisumah and Dhulom sites with 6.06 and 7.06 (Cents/kWh) respectively. This wind turbine (KMW-ERNO) does not give the minimum ECF for the other two sites (Yanbo and Riyadh), where the minimum ECF for these two sites are in using of Acciona6 and GE Energy 2 wind turbines with 5.97 and 12.81 (Cents/kWh) respectively. So, it is recommended to use KMW-ERNO wind turbines in Dhahran, Qaisumah, and Dhulom sites and to use Acciona-6 and GE Energy 2 wind turbines in Yanbo and Riyadh respectively. It is clear from Table 1 and Fig. 14 that Dhahran, Yanbo, and Qaisumah

3.5

3

2.5

2 4.5

5

Capacity factor, CF

shape Parameter,k

0.8 Yanbou Yanbou Dohloum Dohloum Dahhran Dahhran Riyadh Riyadh Qaysoma Qaysoma

5.5

6 6.5 7 Scale Parameter , c

7.5

8

8.5

Fig. 8. The relation between shape parameter and scale parameters for five sites and one hundred wind turbines.

Simulation Results Curve fitting

0.6 0.4 0.2 0 0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

Uav/Ur Fig. 10. The relation between Uav/ur and capacity factor for five sites and one hundred wind turbines under study.

50

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52

35 Yanbou Douhlom Dahhran Riyadh Qaysoma

Energy Price (Cents/kWh)

30

25

20

Table 1 The energy price in Cents/kWh for five sites and best five out of one hundred wind turbines. ECF

Dhahran

Yanbo

Qaisumah

Dhulom

Riyadh

KMW_ERNO Acciona-6 GE Energy 2 Goldwind_3 GE_Energy_3

5.85 6.02 6.06 6.12 6.28

5.98 5.97 6.10 6.06 6.17

6.06 6.36 6.49 6.52 6.63

7.06 7.24 7.34 7.42 7.5

13.11 12.87 12.81 13.32 13.22

15

10

5 0

0.05

0.1

0.15

0.2 0.25 Capacity Factor, CF

0.3

0.35

0.4

Fig. 11. The energy price in Cents/kWh for five sites along with the capacity factor for one hundred wind turbines under study.

Energy Price (Cents/kWh)

30 Simulation Results Curve Fitting

25 20 15 10 5 4.5

Fig. 14. The energy price in Cents/kWh for five sites and best five out of one hundred wind turbines.

5

5.5

6 6.5 7 Average Wind Speed

7.5

8

8.5

Fig. 12. The energy price in Cents/kWh for five sites along with average wind speed for one hundred wind turbines under study.

are the best sites and the highest price for kWh is associated with Riyadh. It is clear that the price of kWh generated in Riyadh is twice its value if we install the wind energy system in Dhahran, Yanbo, or Qaisumah. So, it is not recommended to install wind energy system in Riyadh.

30

Energy Price (Cents/kWh)

Simulation Results Curve Fitting 25 20 15 10 5 0.3

0.35

0.4

0.45

0.5 Uav/Ur

0.55

0.6

0.65

0.7

Fig. 13. The energy price in Cents/kWh for five sites along with (Uav/ur) for one hundred wind turbines under study.

Table 2 The best WTG and ECF value for each site. Site

The best WTG

ECF

Yanbo Dhulom Dhahran Riyadh Qaisumah

Acciona_6 KMW-ERNO KMW-ERNO GE-Energy-2 KMW-ERNO

5.97 7.07 5.85 12.81 6.06

8. Conclusions The calculation of kWh generated from wind energy system is the main factor which determines the visibility of installing wind energy system in any site. Wind turbine can be suitable for one site and may not suitable for the other site. Matching between the site and wind turbine is important before starting installing wind turbines in any country. In this study five sites in Saudi Arabia are used to select the best one and one hundred market available wind turbines. A proposed computer program is introduced in this paper to handle the whole steps of the design of the wind energy system in very short time. Unlimited numbers of sites and wind turbines can be used with this computer program. The salient results from this paper show that the best site from the five sites under study is Dhahran and the suitable wind turbine for this site is KMW-ERNO with 5.85 Cents/kWh. The worst site to install wind energy system is Riyadh with minimum price of kWh of 12.81 Cents/kWh in case of using GE Energy 2 wind turbine. This shows clearly that the price of kWh generated in Riyadh is twice its value if we install the wind energy system in Dhahran. So it is not recommended to install wind energy system in Riyadh. KMW-ERNO wind turbine is the best wind

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52

turbine for Qaisumah and Dhulom sites with 6.06 and 7.06 (Cents/ kWh) respectively. This wind turbine (KMW-ERNO) does not give the minimum ECF for the other two sites (Yanbo and Riyadh), where the minimum ECF for these two sites are in using of Acciona6 and GE Energy 2 wind turbines with 5.97 and 12.81 (Cents/kWh) respectively. So, it is recommended to use KMW-ERNO wind turbines in Dhahran, Qaisumah, and Dhulom sites and to use Acciona-6 and GE Energy 2 wind turbines in Yanbo and Riyadh respectively. Acknowledgments The authors acknowledge the National Plan for Sciences and Technology program (Project No.08-ENE226-02) by King Saud University for the financial support to carry out the research work reported in this paper. Lists of symbols and Abbreviations

DElos, CO&M

the annual electrical transmission loss. the cost of operation and maintenance normalized per unit of energy $ per kWh. the factor relating to the electric losses which occur Klost, t between the wind turbine terminals and the electric grid where the energy is utilized. the factor which depends on how the transmitted wind Kutil, t energy is utilized in the power system. DEper, t the reduction in the annual energy output. the total cost of constructing the facility normalized by Cc rated power $ per kW. s the variance. AAR the average annual return. the annual net energy. ANEt ANWTG average number of WTG. cross sectional area of wind parcel, m2. At the annual utilized energy. AUEt the installed capital cost. CC capacity factor. CF capacity factor. CF the average annual operation and maintenance cost. CO&M COE the cost of energy. the levelized cost of energy. COEL CRF the capital recovery factor. the annual energy production kWh/year. Ea ECF energy cost figure. FCR the fixed charge rate. the annual operation and maintenance cost fraction (of fOM system capital cost). FV the future value. i the general inflation rate. the investment part paid. Ii k the shape parameter. L the lifetime of system. LCC the life-cycle costing. LPC the levelized production cost. N the period of loan, the number of year. NPV the net present value. the cost version of net present value. NPVC the savings version of net present value. NPVs the number of turbines. Nt the power curve for standard conditions. P(u)std the average output power. Pavg the price obtained for electricity $ per kWh. PE the average required load. PLav the rated power. Pr t

51

PV PWF r SP TC ti

the present value. the present worth factor. the discount rate. the simple payback period. the total cost. the years before the start of commercial operation of the wind power installation. TOM the total levelized annual “down line cost”. TPCCS the total price of remote control in central control station. TPMIC the total price of controllers. TPMS the total price of main substation. TPTL the total price of transmission line. TPWTG the total price of wind turbines. the mean wind speed m/s. Uav The ratio of average and rated wind speed. Uav/ur the cut-in speed. uc the cutoff speed. uf the rated speed. ur WTG wind turbine generator. r the corrected monthly air density Kg/m3. References [1] Said SAM, El-Amin IM, Al-Shehri AM. Renewable energy potentials in Saudi Arabia. In: Beirut regional collaboration workshop on energy efficiency and renewable energy technology. American University of Beirut; April 2004. p. 76e82. [2] Rehmana S, Al-Abbadib NM. Wind shear coefficient, turbulence intensity and wind power potential assessment for Dhulom, Saudi Arabia. Renewable Energy 2008;33:2653e60. [3] Alawaji Saleh H. Wind energy resource assessment in Saudi Arabia, part I; network design and description. Renewable Energy International Journal 1996;7(4):319e28. [4] Rehmana S, Halawani TO, Mohandes M. Wind power cost assessment at twenty locations in the kingdom of Saudi Arabia. Renewable Energy 2003;28: pp.573. [5] Rehman S, Ahmad Aftab. Assessment of wind energy potential for coastal locations of the kingdom of Saudi Arabia. Energy 2004;29:pp.1105e15. [6] Rehmana S, El-Aminb IM, Ahmada F, Shaahida SM, Al-Shehrib AM, Bakhashwainb JM. Wind power resource assessment for Rafha, Saudi Arabia. Renewable and Sustainable Energy Reviews 2007:937e50. [7] Al-Abbadi NM. Wind energy resource assessment for five locations in Saudi Arabia. Renewable Energy 2005;30:1489e99. [8] Sahin AZ, Saudi Arabiakal AA. Wind power energy potential at the Northeastern region of Saudi Arabia. Renewable Energy 1998;14(1e4):435e40. [9] Rehmana S, Al-Abbadib NM. Wind power characteristics on the North West Coast of Saudi Arabia. Energy & Environment 2009;20(8). Vol. 21, No. 1. [10] Elhadidy MA, Shaahid SM. Wind resource assessment of Eastern Coastal region of Saudi Arabia. Desalination 2007;209:199e208. [11] EL-Tamaly HH, Hamada M, Eltamaly Ali M. Computer simulation of wind energy system and applications. In: Proceedings Int’l AMSE Conference in system analysis, control & Design July 3e5, 1995;vol. 4. p. 84e93. Brno, Czech Republic. [12] Aspliden CI, Elliot DL, Wendell LL. Resource assessment methods, sitting and performance evaluation. New Jersey: World Scientific; 1986. p. 321e76. [13] Johnson Gary L. Wind energy system. n.j.0.07632, vsa. England Cliffs: Book, Prentic-hall, Inc, energy-the-facts.org/en/home-about-the-project.html; 1985. [14] Lun Isaac YF, Lam Joseph C. A study of Weibull parameters using long-term wind observations. Renewable Energy 2000;20:145e53. [15] Manwell JF, McGowan JG, Rogers AL. Wind energy explained e theory design and application. [16] Kandt A, Brown E, Dominick J, Jurotcih T. Making the economic case for smallscale distributed wind e a screening for distributed generation wind opportunities. In: WindPower Conference, Los Angeles, California June 3e6, 2007. [17] Riggs JL. Engineering economics. New York: MeGraw-Hill; 1982. [18] Kabir Md Ruhul, Rooke Braden, Malinga Dassanayake GD, Fleck Brian A. Comparative life cycle energy, emission, and economic analysis of 100 kW Nameplate wind power generation. Renewable Energy January 2012;37(1): 133e41. [19] Peter Friedman D. Evaluating economic uncertainty of municipal wind turbine projects. Renewable Energy 2010;35:484e9. [20] Nath C. Maintenance cost of wind energy conversion systems. German is cher Lloyd, http//www.germanlloyd.delActivitiesNuTind/public/mainten, html; 1998. [21] Rabl A. Active solar collectors and their applications. Oxford: Oxford University Press; 1985. [22] Karas KC. Wind energy: what does it really cost?. In: Proc. Wind power ’92, AWEA 1992. p. 157e66.

52

A.M. Eltamaly / Renewable Energy 60 (2013) 42e52

[23] EPRI (Electric Power Research Institute) technical assessment guide, vol. 1e3. EPRI; 1979. EPRI report: EPRI-PS-1201 -SR. [24] EPRI (Electric Power Research Institute) technical assessment guide, vol. 1. EPRI; 1989. Rev. 6, EPRI Report: EPRl P-6587-L. [25] Nuclear Energy Agency. The cost of generating electricity in nuclear and coal fired power stations 1983. [26] John Olav Tande. Estimation of cost of energy from wind energy conversion systems. Technical report. Riso National Laboratory & Raymond Hunter; 1994. [27] Troen I, Petersen EL. European wind atlas, ISBN 87-550-1482-8; 1989. p. 656. [28] Derrick A. Development of the measure-correlate-predict strategy for site as assessment. In: Proc. of BWEA 1992. [29] IEA. Power performance testing. In: Frandsen S, Pedersen BM, editors. Expert group study on recommended practices for wind turbine testing and evaluation 1990.

[30] Bartholomy Obadiah Julian. A technical, economic, and environmental assessment of the production of renewable hydrogen from wind in California 2002. [31] ENVIS Centre on Renewable Energy and Environment. Wind energy information 2005/2006. [32] Wind energy e the facts e see [http://www.wind-energy-the-facts.org/en/ home-about-the-project.html]. [33] Jaramillo OA, Borja MA. Wind speed analysis in La Ventosa, Mexico: a bimodal probability distribution case. Renewable Energy 2004;29:1613e30. [34] Yeh Tai-Her, Wang Li. Benefit analysis of wind turbine generators using different economic-cost methods. In: Conf Proc. of Intelligent Systems Applications to Power Systems 2007. p. 1e6. [35] Skidmore EL. Wind erosion climatic erosivity. In Climatic change, vol. 9. D. Reidel Publishing Company; 1986. p. 195e208.