Wind and Energy Projection of the Philippines


The Ateneo Innovation Center is engaged in the research of wind energy resource.  This work is being done in collaboration with Alternergy Philippine Holdings Corporation and Manila Observatory.  Current studies are focused on long-term wind projections for prospecting and next-day wind forecasting.  Long-term wind projections are for profiling the potential wind power production of a wind farm for its entire lifetime while the wind forecasting research is geared towards near-term energy production determination for the spot market.  As a whole, these studies look at the diverse and rich renewable energy potential available of the Philippines as a solution to the energy shortage that is happening in the nation.

Long-term Wind Projections

This study characterizes a wind farm site in Pililla, Rizal for its potential wind power production from 2013 to 2037.  The Regional Climate Model ver. 3 (RegCM3) is used for making the projections.  The data is processed using wind energy assessment methods to determine the wind power density and energy production.
RegCM3 is configured to simulate the Philippine domain at 40km X 40km resolution.  This is downscaled to 10km X 10km resolution in order to attain the spatial scale that is closer to the wind farm region.  The projection incorporates the A1B scenario of the European Centre/Hamburg 5 (ECHAM5) which is the case where economic development continues but mitigation measures for emissions are in place.
To verify the model, the Tanay weather station of the Philippine Atmospheric, Geophysical, and Astronomical Services Administration (PAGASA) is used for comparison.  The map below shows the location of the wind farm and the weather station.  
Since the grid is large enough to cover both the station and wind farm site, it suffices to compare the model results with the Tanay station.  The next figure shows the comparison of wind histograms between the observed and RegCM3 model output:

From the figure above, it shows that the model have some discrepancy for the Northeast Monsoon season (from October to April) but can capture the Southwest Monsoon (from May to September) well. The same is true for the wind directions. The Easterlies is not captured by the model. Nevertheless, the majority of the monsoon season is simulated well by the model. For climate scales, this suffices for the purpose of the study. 

The results are grouped into five-year windows to see the climate trends for the wind power production in the site.  The 2008-2012 years are used as the baseline for the analysis.
The chart above shows the seasonal mean winds of each five-year interval. At this site, the variations are only up to 1m/s which suggest that the power production for this site is stable for its entire lifetime.
It is necessary to look at the power density and capacity factor in assessing power production.  This is to gain insights on the performance of the wind farm.  A Weibull distribution using both unimodal and bimodal are incorporated to the analysis.  The wind power density of the Pililla wind farm is presented below:
The graph presents that there is an increase to the wind power density of the site in comparison to the baseline years.  Since there is minimal difference between the unimodal and bimodal distribution, the unimodal is used for all the analysis.
The average wind speeds and capacity factors are also calculated and the percentage changes are summarized below along with the wind power density.
 Year Ave. Wind Speed   Wind Power Density Capacity Factor 
 2013-2017 4% 6%  8% 
 2018-2022 4%  1%  3% 
 2023-2027 3%  9%  9% 
 2028-2032 5%  12%  12% 
 2033-2037 -1%  1%  0% 
 Average 3%  6%  6% 
The results show that the site will experience an increase in power production in the future.  It will have no gain in the last five years of its lifetime when compared to the baseline.  The increase in power production will decrease the levelized cost of energy for this technology. 
This study has been published in Energy Procedia by Elsevier.  Please go to this site to access the full paper.

Next-day Wind Forecasting

Short-term wind projections is necessary for incorporating wind energy systems into the grid.  This is to ensure the stability of the entire grid in the daily operations.  Merit order is also being integrated into the work since the present energy shortage are due to peak demand periods in the daily utilization of power.  These can be addressed by renewable energy in a cost-effective way in comparison to conventional power plants.  Time series analysis techniques are implemented to determine the expected power production in the next few hours up to 24 hours.
Analysis is done using R and Matlab.  Initial results are computed using ARIMA(7,0,5) and shows that the method can predict the nine hours ahead.  To further improve this, the seasonality of the wind and diurnal characteristics will be incorporated into the analysis.  Other techniques are also being explored like Fourier series analysis.