Climate Exchange

[

] 167

nergy

Methodology

With the assistance of EPC and SMD, rainfall data for Afulilo Dam

and surrounding stations were obtained, along with storage, stream-

flow and power generation records from the Ta’elefaga Powerhouse.

Data requirements for calibrating the hydrological model included:

• Climate (rainfall, temperature and evaporation data)

• Inflow to the dam (catchment streamflow)

• Rainfall-runoff relationship for Afulilo catchment

• Dam release downstream from the dam

• Dam release to Ta’elefaga Powerhouse for power generation

• Dam storage characteristics (full storage/dead storage levels, cross-

section of spillway)

• Storage area-volume-elevation relationship for Afulilo Dam

• Estimated seepage rates and losses

• Turbine efficiency (or rating curves)

• Outflow-to-power-generation relationship at different storage

levels (change in turbine efficiency).

Long-term (1900-2010) daily climate data for Afulilo were generated,

based on Apia climate data and using the stochastic weather generator,

WeatherMan.

Since daily streamflow data were limited for the Afulilo

catchment, an appropriate rainfall-runoff relationship had to be derived

from the streamflow records of the neighbouring Vaipu catchment using

the IHACRES software package.

This relationship was then used to

calculate inflow to Afulilo Dam. Outflow to the Ta’elefaga Powerhouse

was calculated based on daily records of electricity generation fromEPC.

Using the derived inflow and outflow rates, a hydrological water

balance model for Afulilo Dam was developed and calibrated against

historical dam levels as well as power generation data for 1993-2010.

Analysis showed a significant correlation between seasonal storage

volumes and ENSO, with El Niño periods tending to

coincide with lower storage volumes and La Niña periods

tending to coincide with higher ones.

Results

A comparison of observed and modelled storage levels

was made for a validation period of 1 August 1993 to

28 February 2010. While the model will become more

accurate as more hydrological data become available for

EPC, it was considered sufficiently accurate to begin

considering long-term supply and demand scenarios.

Several variables were incorporated into this

ensemble of long-term supply and demand simula-

tions to assess each of their effects on dam storage

and energy generation potential, and to assess dam

performance under different scenarios. These vari-

ables consisted of a matrix of power demand, power

generation capacity, water storage capacity and rain-

fall variability.

The simulations showed that under the current

average power demand, the dam storage level falls

below 10 per cent of usable capacity on an average of

five days per year. If energy demand continues to grow

at the projected 4 per cent per year, then by 2020 the

number of days where usable storage drops below 10

per cent will rise to 70 days per year. The results also

showed a 20 per cent increase in the frequency of low

storage volume under El Niño as compared to La Niña

conditions, further underlining the significant effect of

climate variability on energy security in Samoa.

Capacity improvement options at Afulilo Dam

Source: Australian BoM