Effects of air psychrometrics on the exergetic efficiency of ... - CiteSeerX

Energy 37 (2012) 632e638

Contents lists available at SciVerse ScienceDirect

Energy journal homepage: www.elsevier.com/locate/energy

Effects of air psychrometrics on the exergetic efficiency of a wind farm at a coastal mountainous site e An experimental study G. Xydis* Intelligent Energy Systems Programme, Risø DTU National Laboratory for Sustainable Energy, Frederiksborgvej 399, P.O.B. 49, 4000 Roskilde, Denmark

a r t i c l e i n f o

a b s t r a c t

Article history: Received 29 April 2011 Received in revised form 17 October 2011 Accepted 24 October 2011 Available online 21 November 2011

In this paper, the most important energy and exergy characteristics of wind energy were examined. Atmospheric variables as air temperature, humidity and pressure and their effects on the wind turbine output were investigated toward wind energy exploitation. It was shown that these usually disregarded meteorological parameters while planning new WFs (Wind Farms), in fact, do play an important role in the farm’s overall exergetic efficiency. The wind potential around a coastal mountainous area was studied based on field measurements. Understanding atmospheric parameters variation appears to be of great importance for estimating energy yield in rough terrain and in this paper special focus was given to that. Ó 2011 Elsevier Ltd. All rights reserved.

Keywords: Exergetic efficiency Coastal mountainous areas Wind Atmospheric variables

1. Introduction Accurate wind resource analyses based on measurements e using, most of the times, wind meteorological towers e are absolutely necessary for the exploitation of wind energy. Wind characteristics measured, usually include wind speed with anemometers at different heights, wind direction using wind vanes at different heights and temperature (using thermometers) based on the IEC 61400-12-1 international standard [1]. Greek terrain is mostly mountainous extending many times into the sea as peninsulas. Air flow over mountains or hills, is very advantageous for WFs (Wind Farms) as they tend to increase the wind speed (compared to the incoming flow) because of the obstructions on the incoming wind and therefore are many times preferable as this way the power output is increased (speed-up effect) [2]. In coastal mountainous peninsulas the speed-up effect accompanied with the sea breeze phenomenon, which usually has an effect on the wind resource of coastal areas, have a positive impact on the measured values of the wind speed of the area something which reveals candidate peninsular sites to be chosen as suitable for WF development. The coastal type of these sites makes them also easily accessible with no excessive road works

* Corresponding author. Tel.: þ45 4677 4974, þ45 5180 1554 (mobile); fax: þ45 4677 5688. E-mail addresses: [email protected], [email protected]. 0360-5442/$ e see front matter Ó 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.energy.2011.10.039

needed (and therefore less expensive projects compared to strictly mountainous sites). Atmospheric parameters considered negligible so far and their effect were studied thoroughly on the way they influence exergetic efficiency of WFs. A draft literature review on that, site experimental results and discussion follow in the next sections. 2. Exergy studies on renewable energy sources A large body of literature concerning the applications of exergy analysis has been carried out during the past decades. However, exergy analysis on intermittent renewable energy sources and studies concerning investigations on exergetic efficiency and specially related to wind energy are not that many. Koroneos et al. dealt with many innovative studies for RES [3,4]. However, especially for wind Koroneos et al. [5] dealt with the three kinds of RES in terms of exergetic aspects. In this research the authors concluded for different wind turbines (600 kWe1 MW) that while the wind speed changes between 5 m/s and 9 m/s, the available wind potential for electricity use changes between 35% and 45% due to exergy losses mainly because of the rotor, the gearbox and the generator. S¸ahin et al. [6] estimated mean exergy and energy efficiencies in relation to the wind speed and suggested that exergy efficiency should be used for wind energy sitting in order modeling to be more realistic. Under this concept, Xydis et al. [7] implemented the exergy analysis methodology as a WF sitting tool in Central Peloponnese, Greece, an analysis which showed that gross

G. Xydis / Energy 37 (2012) 632e638

633

Nomenclature A Ci h l L _ m P Pa R Ra RH Rw T U VR

wind turbine rotor swept area (m2) installed capacity of the wind farm (MW) altitude (m) cable length (km) transmission loss (W) air mass flow (kg/s) power load (kW) pressure in the humid air (Pa) resistance (Ohm/km) the gas constant (¼286.9 J/kg K) relative humidity (kg moisture/kg dry air) the water vapor constant (¼ 461.5 J/kg K) air temperature ( C) voltage [kV] wind speed (m/s)

Fig. 1. Two (2) wind masts installed in coastal mountainous sites.

3. Exergetic analysis 3.1. Site experimental results

Greek letters r air density (kg/m3) rda dry air density (kg/m3) u specific humidity or humidity ratio (kg moisture/kg dry air) 4 phase angle between active and reactive power (rad)

Annual Energy Production (AEP) & net AEP may differ significantly based on other parameters variation like transmission losses, air density losses, topographic losses (wake effects) and wind turbine availability. S¸ahin et al. [8] used exergy analysis for each system, applying a point-by-point map analysis giving another approach to wind power systems as exergy maps provided more useful information (compared to energy analysis) regarding losses. Ozgener and Ozgener [9] carried out an exergy and a reliability analysis of a wind turbine proving e among others e and showed that exergy efficiency changes between 0% and 48.7% at different wind speeds, considering pressure differences from the state point. Hepbasli [10] in his important review on exergetic analysis and assessment of renewable energy resources pointed out that differences between energy and exergy efficiencies were proved to be 40% at low wind speeds and up to nearly 55% at high wind speeds. Ozgener et al. [11] investigated exergetic efficiency and various thermoeconomic values of a small wind turbine and as Baskut et al. [12,13] did pointed out the importance of various meteorological parameters with respect to wind speed. Ahmadi and Ehyaei [14] have dealt with an improved approach for exergy analysis of the wind. Based on the same type of installed wind turbine, by varying the cut in rated and “furling” speeds, showed that the energy production can vary a lot while the entropy generation could be decreased up to 76.9%. However, all types of analyses do not take into account the terrain. What would probably be important and of great value is the effect on exergetic efficiency of the terrain combined with the meteorological effects not just to a specific site but to a whole area. Based on Hepbasli’s review [10] exergy is a measure of the maximum useful work that can be done by a system while Van Gool [15] has reported that the maximum improvement in the exergy efficiency for a specific process can be achieved when exergy losses or irreversibilities are minimized. To fill the gap related to wind and exergy analysis aiming at optimizing the generated power by optimizing the sitting of a WF minimizing at the same time exergy losses or irreversibilities in an area, an innovative study has been carried out and is described in this paper.

Wind profile measurements were carried out for specific periods using two (2) meteorological masts on the south of Mt. Helicon and close to the sea (inside Gulf of Corinth) (Fig. 1). Site coordinates, average velocity, period of measurement, height above ground level, and temperature are shown on Table 1. Tools used for elaborating all the measurements and produce estimates of wind speed/energy output (at various distances from the measuring meteorological masts) were WindRose [16] and WAsP (Wind Atlas Analysis and Application Program) [17]. Vector Hellenic Windfarms S.A. operates a certified laboratory (Laboratory of Wind Measurements) from Hellenic Accreditation System S.A. (E.SY.D.) in Greece and the meteorological stations were under the laboratory’s supervision. It is noticeable, by examining the wind roses, that the two main directions (primary and secondary direction) were N and W respectively. This could be explained based on the fact that both sites are located “inside” the Gulf of Corinth as peninsulas and sea breeze affects them. The wind was studied and the same period was chosen from the two (2) sites (Sept. 2006eAug. 2007). In both cases, two (2) 40 m masts were installed made out of steel in tubular form kept in vertical position using tense wires. Anemometers and vanes were placed every 10 m (10; 20; 30; 40). A data logger connected to the available sensors of the mast stored and sent the data to the responsible laboratory using the GSM method. The uncertainty of the measured wind speed for masts “A” and “R” was calculated using the WindRose software [16] at 0.192 and 0.188 m/s respectively.

3.2. Exergy analysis The available output from the proposed WF could be determined based on the flow rate through the rotor (swept area) of the wind turbine. The kinetic energy Ek is:

Table 1 Main measured characteristics of other two (2) sites in the wider area. Site/Code Latitude ( ) A R

Longitude ( )

Mean Period of speed data (ms1) analysis

38.19916 22.944167 6.19 at 40 m. 38.21750 22.908056 6.51 at 40 m.

Height Temperature (magl) ( C)

9 Sept ’06e 307 9 Oct ’07 25 Oct ’05e 404 11 Oct ’07

17.65 16.09

634

Ek ¼

G. Xydis / Energy 37 (2012) 632e638

1 _ R2 ; $m$V 2

(1)

_ are the wind speed and the air mass flow rate where VR and m respectively, and

_ ¼ r$A$VR ; m

(2)

where r is air density, A is the wind turbine rotor swept area equals p$R2.

Thus; Ek ¼

1 $r$p$R2 $VR3 2

(3)

something which means that if the wind speed is known (measured), for a given wind turbine, the kinetic energy can be defined, and since kinetic energy is a form of mechanical energy it can be converted to work unconditionally then the exergy output is also known. Following this concept, a WAsP-based wind resource analysis [17] in the under examination area (Fig. 2) shows the average kinetic energy per unit area (in particular 45 m  45 m) perpendicular to the wind flow measured in [W/m2]. Based on this analysis the energetic potential output of each unit area is revealed with a mean value of 294 W/m2. It can also be seen that an initial planning has been done and 14 locations have been selected for installing wind turbines. The terrain is characterized as complex and the hub heights of the proposed wind turbines to be installed vary from 121 m to 420 m above ground level (m.a.g.l.). However, it should be noted that is assumed, like in most cases, that To and Po were taken as standard-state values, such as 25  C and 1 atm. In practice, the environmental conditions most of the times differ, something which affects eventually real electricity production for WFs, but it has not been taken into consideration so far. What should be determined, in order to specify the exergy efficiency is the exergy destruction (losses).

It is widely accepted e on coastal areas e that air is nothing else than a mixture of dry air and water vapor. Therefore, as dry air is more dense that humid air and since the under examination area for WF development is also mountainous (basically hilly, with an average elevation of 200 m), exergy losses because of air density r are even higher. Air density based on [6,12,13] can be expressed as:

r ¼ ðPa=Ra$TÞ$ð1 þ uÞ=ð1 þ u$Rw=RaÞ

(4)

where Ra (¼286.9 J/kg K) the gas constant air, Rw (¼461.5 J/kg K) the water vapor constant, u the specific humidity or humidity ratio (in kg/kg), and Pa the pressure in the humid air (in Pa). The dry air density rda can be mathematically represented as:

rda ¼ Pa=Ra$T

(5)

where T the air temperature and since

Rw=Ra ¼ 1:609

(6)

and therefore, the final expression of r becomes:

r ¼ rda $ð1 þ uÞ=ð1 þ 1:609$uÞ

(7)

Therefore, estimating the humidity ratio and the dry air density (equation (5) includes temperature and pressure), actual air density can be found and exergy losses due to “air density variation” can be determined. Based on [18,19], air pressure is related to altitude using the following formula:

h ¼ 44:3308  4946:54$Pa0:190263 ;

(8)

where h is the altitude (in meters) and Pa (in Pa). Since we already know the height of the proposed wind turbines in the project we can calculate the Pa which is needed for finding r. Knowing the average air temperature (considered as dry-bulb temperature) and the relative humidity (RH) measured from

Fig. 2. Power per unit area perpendicular to the wind flow in W/m2 and proposed wind farm planning.

G. Xydis / Energy 37 (2012) 632e638

635

Table 2 Calculation of air density r for all heights for each month in [kg/m3]

r

H ¼ 400

H ¼ 381

H ¼ 420

H ¼ 220

H ¼ 240

H ¼ 282

H ¼ 301

H ¼ 190

H ¼ 121

H ¼ 280

H ¼ 268

Sept.2006 Oct.2006 Nov. 2006 Dec. 2006 Jan. 2007 Feb. 2007 Mar. 2007 Apr. 2007 May 2007 Jun. 2007 Jul. 2007 Aug. 2007 Average

1.136 1.154 1.173 1.184 1.180 1.188 1.179 1.165 1.143 1.122 1.111 1.115 1.154

1.139 1.157 1.175 1.187 1.183 1.191 1.182 1.167 1.146 1.125 1.114 1.117 1.157

1.134 1.152 1.170 1.182 1.177 1.185 1.176 1.162 1.141 1.120 1.108 1.112 1.151

1.161 1.179 1.198 1.210 1.206 1.214 1.205 1.190 1.168 1.147 1.135 1.139 1.179

1.158 1.177 1.195 1.207 1.203 1.211 1.202 1.187 1.165 1.144 1.132 1.136 1.177

1.153 1.171 1.189 1.201 1.197 1.205 1.196 1.181 1.160 1.138 1.127 1.130 1.171

1.150 1.168 1.186 1.198 1.194 1.202 1.193 1.179 1.157 1.136 1.124 1.128 1.168

1.165 1.184 1.202 1.214 1.210 1.218 1.209 1.194 1.172 1.151 1.139 1.143 1.184

1.175 1.193 1.212 1.224 1.220 1.228 1.219 1.204 1.182 1.160 1.149 1.152 1.193

1.153 1.171 1.189 1.201 1.197 1.205 1.196 1.182 1.160 1.138 1.127 1.131 1.171

1.154 1.173 1.191 1.203 1.199 1.207 1.198 1.183 1.162 1.140 1.129 1.132 1.173

“Elefsina” meteorological station, a coastal site as well which is relatively close (not more than 40 km away) [20], based on air psychrometrics humidity ratio u is determined by the psychrometric chart. A psychrometric chart is a graph that describes graphically the physical and thermodynamic properties of moist air and how these are related to each other. The properties found on a psychrometric chart are the DBT (dry-bulb temperature), WBT (wet-bulb temperature), RH (relative humidity), humidity ratio, specific enthalpy, and specific volume [21]. 4. Discussion Since u is identified, using the equations (5)e(8) seasonal air density r can be determined (Table 2). Based on that, the exergy losses because of air density can be estimated for every different wind turbine location in the under examination area. It is rather obvious that air density varies not only from site to site but also seasonally (1.108e1.228 kg/m3) (Fig. 3). This has a significant effect on the final production from the proposed WF. It is seen that in general follow the same trend. Based on the yearly analysis of the wind speed, air density (at 400 m), humidity ratio and air temperature in the area (Fig. 4), what can be seen at once is the dominating relation of r and T (predictable however from the equation (4)). It can also be seen and understood that the lower the air temperature and humidity ratio are, the higher the air density (at least from autumn till spring) is. Following this thinking, the higher the air density is the lower the exergy losses are due to psychrometric variables. It is obvious therefore, that during the winter time the average wind speed is higher.

As mentioned, exergy analysis can be used as a tool to measure and evaluate interconnected WFs considering their losses (topographic & wake losses, cable losses, transformer or substation losses, technical availability losses, and air density losses) calculating the maximum useful work that can be derived from a WF and not only estimating the kinetic energy of the fluid and the maximum work extracted from it in the optimum conditions. In the research implemented and presented in this paper the focus was on identifying the losses due to seasonal variation of the air density and exergetically find the effect of it in the net production of a proposed WF. Therefore, calculated air density losses were inserted to the model from and with the help of the Wind Turbine Power Calculator of Danish Wind Industry Association [22] and the Swiss Wind Power Data website [23] an updated power map is produced which includes not only the topographical and wake effects as usual, but the losses due to air density variation (Fig. 5). Based on this resource grid analysis, the produced power map gives the ability to the wind developer not only to optimize the planning of the WF taking into consideration the topographical and wake effects but also to take into account the air density losses which is not usual so far. Adding up also the electrical losses (internal interconnection medium voltage losses and transformer losses) and the wind turbine’s technical availability losses, a fixed percentage for the proposed WF it is possible to get the exergetic efficiency map on Fig. 7. Exergy flow diagram (Sankey diagrams) shows proportionally the exergy flow quantity and therefore the overall losses of the proposed WF (Fig. 6). Following the analysis of [24] the electrical losses are taken into account the cable transmission losses based on the equation

Fig. 3. Air density seasonal variation of the highest and lowest proposed site.

636

G. Xydis / Energy 37 (2012) 632e638

Fig. 4. 12-month comparative analysis of wind speed, air density, humidity ratio and air temperature.

Fig. 5. Topographical and wake effects and air density variation losses.

G. Xydis / Energy 37 (2012) 632e638

637

Fig. 6. Sankey flow diagram describing the losses of the proposed WF.

L ¼ k$P 2 ½W;

(9)

where

k ¼

 R$l  2 f $ 1 þ tan U2

(10)

L is the transmission loss [W] along the cable segment, P is the power load [kW]. R represents resistance in [Ohm/km], l cable

length [km], 4 phase angle [rad] between active and reactive power and U the voltage level [kV]. The electrical losses were estimated taking into account the fact that the WF is planned to be 35 MW, and therefore the overall electrical losses will be specified from the medium voltage losses for the interconnection of the wind turbines and the power station distribution transformer 20/150 kV losses. Based on Eqs. (9) and (10) and the initial planning (Fig. 2) and the medium voltage cable route the cable losses were calculated at 2.45%. Adding up the average losses for each Wind Turbine of the

Fig. 7. Power per unit area perpendicular to the wind flow in W/m2 including exergy losses due to air density variation.

638

G. Xydis / Energy 37 (2012) 632e638

internal Low Voltage/Medium Voltage (LV/MV) transformer (0.6%) and the WF substation Medium Voltage/High Voltage (MV/HV) (0.3%) [25], the sum of the electrical losses are 3.35% for the proposed WF. Following the approach of [7], exergy efficiency of the proposed WF, including all losses can be estimated by using the equation

Exergy Efficiency ¼

NetAEP $100%; 8760$Ci

(11)

where Net AEP is the Net Energy [MWh] produced, 8760 h are the total hours within a year (365 days  24 hours), and Ci the installed capacity of the WF [MW]. An updated resource grid analysis in the frames of power density shows that in the area under examination shows that power density varies from 108 to 716 W/m2. The outcome including all foreseen losses is a more accurate exergetic power density map than the usual results from typical resource grid analyses using wind analysis software (Fig. 7). 5. Conclusions In this paper the effects on exergetic efficiency of a proposed WF based on the psychrometric variables variation as air temperature, pressure, humidity ratio through air density were examined. Energy and exergy efficiency power density maps were introduced to provide a common basis for project developers and to point out parameters neglected so far. The results showed that on average e based on the proposed planning e almost 6% are the air density losses. The air psychrometric analysis showed that the air density varies not only from site to site but also on a monthly basis (1.108e1.228 kg/m3). This variation effects WF productivity. From November to April (lower temperatures) air density is higher (wind speed is also higher) while humidity ratio is low. A graphical representation showed the direct relation of these four parameters for each day of the year. It was proved that the seasonal variation of the air density plays a more important role in the determination of the WF losses than the wake or topographical effects. In specific the losses because of air density were 1.4 times more than wake and topographical losses. It should be noted that this is happening in an area with average altitude of 300 m. In areas with higher altitude this site, the air density losses will be even greater especially during summer months when humidity ratio is higher than winter. This is more noticed in coastal areas e as in this case e as the level of humidity in the air is usually higher. All incorporated losses to the system and an a new resource grid analysis produced as an output an exergetic power density map which could be used from wind project developers for a more precise and accurate prediction of wind energy production. Acknowledgments The preparation of this paper would not have been possible without the support of the Certified Laboratory of Wind

Measurements of Vector Hellenic Windfarms S.A. The used data array was made up of the average (1 value per second e 600 values in 10 min) and maximum 10-min wind speed values in the sites monitored by the laboratory of Wind Measurements of Vector e Hellenic Wind Farms S.A. References [1] Windturbines-Part12-1: power performance measurements of electricity producing wind turbines (IEC 61400-12-1 International standard). IEC; 2005. [2] Røkenes K, Krogstad P-A. Wind tunnel simulation of terrain effects on wind farm siting. Wind Energy 2009;12(4):391e410. [3] Koroneos C, Dompros A, Roumbas G. Renewable energy driven desalination systems modelling. Journal of Cleaner Production 2007;15(5):449e64. [4] Koroneos C, Dompros A, Roumbas G. Hydrogen production via biomass gasification e A life cycle assessment approach. Chemical Engineering and Processing: Process Intensification 2008;47(8):1267e74. [5] Koroneos C, Spachos T, Moussiopoulos N. Exergy analysis of renewable energy sources. Renewable Energy 2003;28(2):295e310. [6] S¸ahin AD, Dincer I, Rosen MA. Thermodynamic analysis of wind energy. International Journal of Energy Research 2006;30(8):553e66. [7] Xydis G, Koroneos C, Loizidou M. Exergy analysis in a wind speed prognostic model as a wind farm sitting selection tool: a case study in Southern Greece. Applied Energy 2009;86(11):2411e20. [8] S ̧ahin AD, Dincer I, Rosen MA. New spatio-temporal wind exergy maps. Journal of Energy Resources Technology, Transactions of the ASME 2006; 128(3):194e201. [9] Ozgener O, Ozgener L. Exergy and reliability analysis of wind turbine systems: a case study. Renewable and Sustainable Energy Reviews 2007;11(8): 1811e26. [10] Hepbasli A. A key review on exergetic analysis and assessment of renewable energy resources for a sustainable future. Renewable and Sustainable Energy Reviews 2008;12(3):593e661. [11] Ozgener Onder, Ozgener Leyla, Dincer Ibrahim. Analysis of some exergoeconomic parameters of small wind turbine system. International Journal of Green Energy 2009;6(1):42e56. [12] Baskut O, Ozgener O, Ozgener L. Effects of meteorological variables on exergetic efficiency of wind turbine power plants. Renewable and Sustainable Energy Reviews 2010;14(9):3237e41. [13] Baskut O, Ozgener O, Ozgener L. Second law analysis of wind turbine power plants: Cesme, Izmir example. Energy 2011;36(5):2535e42. [14] Ahmadi A, Ehyaei MA. Exergy analysis of a wind turbine. International Journal of Exergy 2009;6(4):457e76. [15] Van Gool W. Energy policy: fairly tales and factualities. In: Soares ODD, Martins da Cruz A, Costa Pereira G, Soares IMRT, Reis AJPS, editors. Innovation and technology-strategies and policies. Dordrecht: Kluwer; 1997. p. 93e105. [16] WindRose e A wind data analysis tool, user’s guide. Centre for Renewable Energy Sources. Available from: http://www.windrose.gr; 2010. [17] Mortensen NG, Landsberg L, Troen I, Petersen EL. Wind Atlas Analysis and Application Program (WAsP). Roskilde, Denmark: Risø Nat. Labs; 1993. p. 126. [18] International Organization for Standardization. Standard atmosphere. ISO 2533; 1975. 1975. [19] A quick derivation relating altitude to air pressure. Portland State Aerospace Society; 2004. [20] Climatology data, Hellenic National Meteorological Service, 2011 Available from: http://www.hnms.gr/hnms/english/index_html? [21] Zhang Z, Pate M. A methodology for implementing a psychrometric chart in a computer graphics system. ASHRAE Transactions 1988;94:1. [22] Wind turbine power calculator, Danish Wind Industry Association, 2011 Available from: http://www.vindselskab.dk/en/tour/wres/pow/index.htm [23] The Swiss wind power data website, power calculator, 2011 Available from: http://www.wind-data.ch/tools/powercalc.php?lng¼en [24] Klaus-Ole Vogstad. Energy resource planning; integrating wind power. Diploma thesis, SMU/pav. B/NTNU/7034 Trondheim/Norway [25] Use and maintenance of oil-immersed distribution transformers, Schneider Electric, Available from:http://www.domaxinternational.com/images/Schneider %20Distribution%20Transformer1.pdf.