Sei sulla pagina 1di 3

EE5003: Energy Systems

June 2015
Tutorial
Energy Demand and Demand Forecasting

(a) A trend analysis performed on electricity sales and generation of a developing country
gave the following results from a least-square analysis.

Ln ( S ) 114 . 36 0. 0626 Y

Sum of residuals, R 2 =0.988, t-value=-23.6 for the intercept and 25.7 for the
coefficient. S is the sales in GWh. Y is the year (such as 2012, 2013, ...)

Ln ( G ) 124 .11 0. 0675 Y

Sum of residuals, R 2 =0.982, t-value=-19.2 for the intercept and 20.8 for the
coefficient. G is the generation in GWh. Y is the year (such as 2012, 2013 ...)

(i) Comment on the statistical significance of time as a determinant of electricity demand


in the country.
(ii) The annual growth rate of sales in year t is defined as,

S t S t 1
gt 100%
S t 1

Calculate the annual growth rates for sales and generation indicated by the results of
the statistical analysis, for years 2012, 2013 and 2014. Also comment on the losses in
the network.

(iii) The demand growth for electricity is expected to slow-down from 2016 onwards, and
is expected to exhibit signs of saturation. If the sales are expected to reach 195,000
GWh per year by the year 2030, develop a model of the form,

S 2 k1 ( t k 2 )

to forecast the sales in the period 2015-2020. k 1 and k 2 are constants to be derived. t
is the year (such as 2013,2014, ...). S is the sales in GWh. Results of historic trend
analysis may be used to calculate the sales in the year 2012.

Using the forecast model, estimate the sales in the year 2016.

(b) Explain the limitations of trend extrapolation in demand forecasting, and describe how
the econometric method and the end use method overcome these limitations.

ts/345324681.doc/February 13, 2017


Q2.
(a)The electricity intensity E i in the national economy is defined as,
Y
Ei
S
where S is the electricity sold per person per year, and Y is the national gross
domestic product (GDP) per person per year.

Table Q2 shows the historic variation of sales and GDP in a developing country.
Estimate the electricity intensity in each year during 1986-1995, and comment on
the result.
Year 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
GDP(millions of 114,261 115,922 119,050 121,729 129,244 135,204 140,990 150,783 157,115 165,286

constant value
currency)
Electricity 2,232
Sold 2,253 2,371 2,353 2,608 2,742 2,916 3,270 3,565 3,886
(GWh)
Table Q2 (a)

(b) Self and cross-price elasticity of demand for diesel fuel in a country, is shown below:-

Price change
Electricity Gasoline Diesel LPG
Diesel demand 0.4 1.0 -0.2 0.0
change
Table Q2 (b)

(i) Identify the possible reasons for the above figures and their magnitudes, including
ability for substitution.
(ii) Estimate the change in demand for diesel, if the prices of all the above energy
products are increased by 10%.

Q3

(a) The specific value addition for electricity E in the national economy is defined as,
Y
E
S
where S is the electricity sold per person per year, and Y is the gross domestic
product (GDP) per person per year.

Table Q3(a) shows the recent variations of electricity sales and GDP in a developing
country. Calculate the per capita economic output and electricity use for each year.

2
Estimate the specific value addition for electricity in each year during 2006-2010, and
comment on the result.

Year 2006 2007 2008 2009 2010


GDP(millions of constant value 135,20 140,99 150,78 157,115 165,28
currency) 4 0 3 6
Electricity Sold (GWh) 7,766 8,169 8,350 8,372 9,191
Mid-year population (million) 19.88 20.01 20.21 20.45
20.653
6 0 7 0
Table Q3(a)

(b) Gasoline sales in Sri Lanka for ten years, is shown in Table Q3(b).

Year 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Gasoline
sales 249.5 286.1 337.9 350.3 463.0 471.1 509.1 510.6 539.6 616.5
(thousan 1 0 3 7 0 6 0 4 4 5
d tonne)

A trend analysis model of the form,

Ln Dg 5.507 0.0957 Y 2000 ---------------- (1)

where Dg = demand for gasoline in thousand tonne


Y = year (such as 2001, 2002)

was derived using the above data, and the sum of residuals was calculated to be, R 2
= 0.936.

(i) Comment on the statistical significance of time as a determinant of gasoline demand


in the country.
(ii) The annual growth rate of gasoline sales in year t is defined as,

Dgt Dgt 1
gt 100%
Dgt 1
Calculate the growth rate represented by equation (1) above.

(iii) Using the model in equation (1), estimate the demand for gasoline in year 2014.

(iv) List other forecasting techniques that could be used to forecast gasoline sales, along
with their strengths and weaknesses.

Potrebbero piacerti anche