Preliminary study of long-term wind characteristics of the Mexican Yucatán Peninsula

Abstract

Mexico’s Yucatán Peninsula is one of the most promising areas for wind energy development within the Latin American region but no comprehensive assessment of wind resource has been previously published. This research presents a preliminary analysis of the meteorological parameters relevant to the wind resource in order to find patterns in their long-term behaviour and to establish a foundation for subsequent research into the wind power potential of the Yucatán Peninsula. Three meteorological stations with data measured for a period between 10 and 20 years were used in this study. The monthly trends of ambient temperature, atmospheric pressure and wind speed data were identified and are discussed. The directional behaviour of the winds, their frequency distributions and the related Weibull parameters are presented. Wind power densities for the study sites have been estimated and have been shown to be relatively low (wind power class 1), though a larger number of suitable sites needs to be studied before a definitive resource evaluation can be reported.

Introduction

Wind power is seen as one of the most effective means available to combat the twin crises of global climate change and energy security, providing a possible solution to the problems associated with volatility in the fossil fuel markets for coal, gas and oil. The “Global Wind Energy Council” (GWEC) in their report “Global Wind Report 2006” [1] presented the state of development of wind energy for leading countries in each geographical region for the period 2000–2006 and a global forecast for 2007–2010 of the cumulative capacity of wind energy installations, which is expected to reach 149.5 GW by 2010. According to this report, the annual market for wind energy grew at a rate of 32% in 2006, with over 15,000 MW of new capacity installed worldwide. The market continued to broaden with installations in over 70 countries, establishing wind power as the leading renewable energy technology. Globally, the value of new generating plant installed in 2006 reached US$24 billion. While Europe continues as the world leader, with 65% of the global market, the United States was the leader in terms of new installed capacity in 2006 with 2500 MW of new installed plant. On the other hand, new capacity installed in Latin America and the Caribbean during 2006 was 296 MW, increasing the total installed capacity to 508 MW. Mexico is one of the most promising areas for wind energy development being in the list of countries that more than doubled its installed capacity with 87 MW mainly located in the Isthmus of Tehuantepec in the State of Oaxaca.

A basic requirement when developing wind power projects is to study the geographical distribution of wind and its main characteristic parameters. Herberta et al. [2] presented in 2007 a general review of wind energy technologies briefly describing the main results published in the scientific literature, covering more than 20 research projects undertaken around the world. This review was particularly concerned with wind resource assessment. Average wind speed and temporal wind patterns have been identified using short-term measurements, such as in the study by Essa and Embaby [3] to describe the winds of a site close to the Egyptian Mediterranean coast over one year. Similarly, Li and Li [4] analyzed data over 5 years to determine the annual, seasonal and diurnal wind characteristics for the Waterloo region in Canada.

Long-term studies for periods of at least 10 years have been reported using wind data measured at meteorological stations on islands and in coastal areas. Farrugia and Scerri [5], collated wind data over 24 years at a height of 11 m above ground level (a.g.l.) from Luqa International Airport in the Maltese Archipelago to analyze wind parameters and to identify diurnal trends. Shata and Hanitsch [6], [7], presented two complementary studies, where the wind power potential of two different Egyptian coastal regions were analyzed in terms of their wind characteristics over 10 years at a height of 10 m a.g.l. using monthly averaged measurements for a total of 17 coastal meteorological stations.

Another Mediterranean coastal study, in the eastern region of Turkey, was undertaken by Sahin et al. [8] with the explicit purpose of “quantify[ing] the wind energy potential of and identify locations with better wind resources”. Seven meteorological stations were chosen in this study at sites situated between 4 m and 100 m above mean sea level (a.m.s.l.) with hourly measurements over 10 years (1992–2001). The Weibull parameters were computed for all sites using the last 5 years of the study period. A diurnal pattern was identified with the lowest wind speeds around 7:00 am local time and the highest wind speeds during the afternoon. The monthly mean wind speeds showed a seasonal pattern with the highest wind speeds in July and August and the lowest in November and December, in almost all of the stations that were studied.

Long-term studies for peninsular regions have been also undertaken. For a site located on the west coast of Saudi Arabia, Rehman [9], studied the annual, seasonal and diurnal patterns of the wind speed over 14 years in order to make wind energy yield calculations using wind turbines of different capacities. Because of the relatively low wind speeds identified, the author found better capacities factors for smaller wind turbines. Later, Rehman and Ahmad [10] extended their study to five meteorological stations in coastal sites around Saudi Arabia. In this study, long-term hourly mean wind speeds were estimated at three different heights by means of the power law and the authors presented the hourly mean wind speeds during an average day and the monthly mean wind speeds during an average year at four different heights for each site.

In tropical regions, Tchinda and Kaptouom [11] reported in 2003 ra study of the wind energy distributions for two Cameroon provinces using 10 years of meteorological data. The authors found that the annual mean wind speeds for the windiest site exceeded 2 m/s for over 53% of the time and that the wind resource could be useful for applications such as small distributed water pumping systems. Further studies to reinforce the results presented were proposed.

In the case of Mexico, Jaramillo et al. [12] in 2004 undertook a study of the wind potential along the coast line of “Baja California Sur” in the North-West of México. Measurement masts at 15 sites were installed with instruments at 10 m a.g.l. and a year of data was collected (February 1997–1998) by the Electrical Research Institute. After a preliminary evaluation of all 15 sites, the authors chose one site to study the monthly wind speed behaviour, calculating the wind rose, the Weibull distribution and the frequency distribution.

Many of the studies mentioned above used a Weibull probability distribution function (PDF) to represent the statistic behaviour of the wind. Nevertheless, other distributions have been used for this purpose. Akpinar and Akpinar reported initially used a Weibull PDF to characterize the wind energy potential in Turkey [13] but later, in 2007, extended the study reporting better results using the Maximum Entropy Principle (MEP) when assessing the wind energy potential for the range of study sites [14]. More recently, in 2008, Vicente [15], published the results of a comprehensive study undertaken in the Canary Islands, which shows that the use of probability distribution functions more complex than the Weibull distribution to fit wind speed experimental data did not give a significant improvement in the prediction of annual mean energy yield, particularly in cases where the probability of null wind speed was close to zero.

The studies described above have been undertaken mainly in last 10 years in different regions around the world close to coastal sites. The majority of these studies used wind data measured at meteorological station at 10 m a.g.l. during long-term periods to compute wind averages, identify trends over different time scales and to analyze the wind energy potential at the study sites. To date, no proper assessment of the wind resource in the Yucatán Peninsula of México has been published. Therefore, the Autonomous University of Yucatán, México, in collaboration with the Centre for Renewable Energy System Technology (CREST) at Loughborough University, UK, have undertaken a study to assess the wind power potential of the Yucatán Peninsula following the general methodology used in the studies reviewed above. Firstly, three coastal sites were selected as a result of a survey conducted to locate available data recorded by the Observatories of the “Servicio Meteorológico Nacional de la Comisión Nacional del Agua” (National Meteorological Service). Secondly, a conventional procedure to validate the data measured over a period between 10 and 20 years was applied for each study site. Then, the monthly and yearly behaviours of the main environmental parameters were studied. Finally, wind speed data were analyzed to assess temporal patterns as well as the wind direction statistics, wind frequency distributions and estimated power densities.

The main purpose of this research is to present, for the first time, a preliminary analysis of the long-term behaviour of meteorological parameters relevant to the wind resource in the Yucatán Peninsula.