Energy and Power Engineering, 2013, 5, 889-896
doi:10.4236/epe.2013.54B171 Published Online July 2013 (
Optimal Scheme with Load Forecasting for Demand Side
Management (DSM) in Residential Areas
Mohamed AboGaleela, Magdy El-Marsafawy, Mohamed El-Sobki
Electrical power and Machines department, Faculty of engineering Cairo University, Giza Egypt
Received February, 2013
Utilities around the world have been considering Demand Side Management (DSM) in their strategic planning. The
costs of constructing and operating a new capacity generation unit are increasing everyday as well as Transmission and
distribution and land issues for new generation plants, which force the utilities to search for another alternatives without
any additional constraints on customers comfort level or quality of delivered product. De can be defined as the selection,
planning, and implementation of measures intended to have an influence on the demand or customer-side of the electric
meter, either caused directly or stimulated indirectly by the utility. DSM programs are peak clipping, Valley filling,
Load shifting, Load building, energy conservation and flexible load shape. The main Target of this paper is to show the
relation between DSM and Load Forecasting. Moreover, it highlights on the effect of applying DSM on Forecasted de-
mands and how this affects the planning strategies for utility companies. This target will be clearly illustrated through
applying the developed algorithm in this paper on an existing residential compound in Cairo-Egypt.
Keywords: Component; Demand Side Management(DSM); Load factor(L.F.); Short Term Load Forecatsing(STLF);
Long Term Load Forecasting(LTLF); Artificial Neural Network(ANN)
1. Introduction
The most common rationale for Demand Side Manage-
ment in the Power Sector is that it is often more cost ef-
fective and socially beneficial to manage electricity de-
mand through investment in efficiency and other demand
side measures than to increase power supply or transmis-
sion capacity. DSM programs are used to eliminate or
reduce the need for additional peak or base load generat-
ing capacity and/or distribution facilities. DSM also per-
mits existing generation to meet the needs of a larger
number of consumers or defers or reduces the need for
new capacity.
Utilities, however, can benefit from these reductions or
shifts in customer energy use. For some utilities, DSM
programs can help them reduce their peak power pur-
chases on the wholesale market, thereby lowering their
overall cost of operations. In the short term, DSM pro-
grams can reduce energy costs for utilities, and in the
long term, DSM programs can help limit the need for
utilities to build new power plants, distribution, and trans-
mission lines. In short, a DSM program can be much
cheaper to implement than building a new generating plant.
2. DSM Programs
DSM Program is a strategy used to control the load pro-
file indirectly in order to achieve the utility objectives.
These objectives are [1]:
To have the load factor as close as possible to 1.0.
To have the peak load within the system capacity.
By achieving the previous objectives, the utility would
get the maximum possible energy from the installed units,
thus maximizing the total profit and minimizing the av-
erage cost per kWh.
Common techniques used for load shaping are peak
clipping, valley filling, load shifting, strategic conserva-
tion, strategic load growth, and flexible load shape as
shown in Figure 1 [2].
DSM program can also be classified into [3]:
EE (Energy efficiency) programs which are de-
signed to reduce electricity consumption throughout the
year by focusing on reducing energy consumption and
overall energy demand.
DR (Demand Response) programs which are auto-
matic with a processing unit having the right to moderate
or turn-off certain appliances (e.g. air-conditioners, pool
pumps, washing machines, etc.) for a short period of time
at customer sites.
3. Load Forecasting
Electric load forecasting is an important aspect in elec-
Copyright © 2013 SciRes. EPE
trical power industry. It is important to determine the
future demand for power as far in advance as possible.
According to the foreseen load the company makes in-
vestments and decisions on buying energy from the gen-
erating companies, and planning for maintenance and
expansion. It is therefore absolutely necessary to have
some knowledge of future power consumption. Electric
power distributors require a tool that allows them to pre-
dict the load in order to support its management and
make more efficient in planning formulation. Accurate
prediction of electric load is difficult. A large number of
the classical prediction models are inappropriate for this
modeling because of their requirement [4].
The prediction of the electric load at a future time is a
challenging problem because of the diverse characteris-
tics of the electrical load and the uncertainly associated
with them. A typical daily variation of electric loads is
shown in Figure 2. The characteristics of the electric
load depend on the nature of the users and the end use
devices utilized such as motors, air conditioners, lighting,
etc. From this point of view the electric load can be sepa-
rated into four major categories [5].
In a few words load forecasting is very important now
days for the utility companies especially those using the
DSM alternatives, as DSM alternatives will be applied on
the forecasted load demand instead of the existed load.
Importance of load forecasting in the deregulated
market is considered as follows:
Purchasing, generation, sales
Area planning
Infrastructure development/capital expenditure de-
cision making
Load Forecasting is divided into three main Types:
short-term load forecasting (STLF)
medium-term load forecasting (MTLF)
long-term load forecasting (LTLF)
Load Forecasting Methods
Over the last few decades a number of forecasting
methods have been developed. The different forecasting
methods could be stated as follows [6]:
Regression Models.
Time serried approaches.
Similar day approach.
Artificial Neural Networks (ANN).
ANN is used as the forecasting tool presented in this
Figure 1. Load shape objectives.
Figure 2. Typical daily load variation.
Copyright © 2013 SciRes. EPE
4. Simulation Tools
In this paper three tools are introduced:
4.1. Load Forecasting Tool (ANN)
The use of artificial neural networks (ANN or simply NN)
has been a widely studied electric load forecasting tech-
nique since1990 [7] Neural networks are essentially
non-linear circuits that have the demonstrated capability
to do non-linear curve fitting. The outputs of an artificial
neural network are some linear or nonlinear mathemati-
cal function of its inputs. The inputs may be the outputs
of other network elements as well as actual network in-
puts. In practice network elements are arranged in a rela-
tively small number of connected layers of elements be-
tween network inputs and outputs. Feedback paths are
sometimes used.
ANN used is MATLAB based.
4.2. Smart Meter Tool
A smart tool named-SEP2 Manager/ISKRA- is used
through this work having the following features:
Hour by Hour power consumption is recorded
through the meter.
Generates the data to Excel sheets.
Plotting the overall daily load curve for the whole
Figure 3. MATLAB ANN tool main page.
Figure 4. Main page of Smart meter tool[9].
Copyright © 2013 SciRes. EPE
Figure 5. Generated Data tables screen[9].
Figure 6. Output load curves from the tool[9].
Figure 7. Solver Main Screen.
Copyright © 2013 SciRes. EPE
Copyright © 2013 SciRes. EPE
4.3. DSM Optimization Tool (Solver-Excel)
ith Algorithm
Outate development located in New Cai-
introduced in order
The DSM solution tool used is SOLVER EXCEL
used for optimization. The main screen of solver is
shown in figure 6. The objective function is written in
the target cells and the constraints are written in the
changing cells.
5. Practical Case Study w
r case is Real es
ro, Egypt which is fed by electric power on the22 kV
with contracted capacity till 2012 is 20 MW. This com-
pound is fed through a private sector company.
In this section load forecasting is applied using ANN
on the previous case study. Where the day of the maxi-
mum demand for the previous compound was in the first
week of August, so forecasting is applied to estimate the
demand of this day in the next previous three years (2013,
2014&2014).Then DSM load shifting is applied on the
forecasted demand.
The Algorithm steps could be summarized as follows:
The Historical data of the load is obtained from the
The historical data for the weather is obtained from
the weather forecasting Agencies.
These data are inserted to the ANN toolbox based
under MATLAB.
A forecasting model is developed based on the data
entry to the ANN toolbox.
A validation step should be made to the developed
forecasting model.
The model generates the new load demand at cer-
tain required days through entering the weather condi-
tions at these years..
Finally the new loads curves are plotted through
DSM is Load shifting is applied on the new fore-
casted load curves.
Problem Formulation for DS M
A DSM optimization algorithm is
shift loads for achieving the following objective func-
tion which is maximizing load factor of the system with
reducing residual Area between load and supply (i.e.
increasing shared area between supply and load) and this
DSM will be applied in the forecasted load curves that its
load curve exceeds the supply curve in some regions [8].
Objective function:
ii 1
N jjJ
pij tj
L.F.: is the system load factor.
,j)) : is the demand of load type i at time interval
number j.
N: is the total number of load demand types.
J: is the total number of time intervals.
PTO (j): is the total demand for all the loads typ
j. from j=1 to j=J over the time interval number
K: is the number of time interval at which the
maximum demand for all the load types numbers from
i=1, N over all the time duration from j = 1, J occurs.
PTO (K): is the maximum of total demand for all
the loads types.
The previous objective function is subjected to the
following constraints:
Equality constraint: Energy consumption is the
same before and after shifting.
 
,* ,*()
 
jJ jJ
iN iN
ij i
pnewi jtjpoldijt j
Inequality constraint: Difference between supply
and load should be positive.
()sj PTOkzero
Inequality constraint: Peak demand after applying
shifting algorithm should be less than at before shift-
Pnew (i,j) is the new power of each load type(i) at
rval(j) after applying shifting algorithm.
tion stage for the forecasting
tput demand profiles
__pmax oldpmax new
time inte
Pold (i,j) is the old power of each load type(i) at
time interval(j) before applying shifting algorithm.
PTO (j): is the total load power at time interval (j).
S (j): is the supply value at time interval j.
6. Simulation Results
6.1. Load Forecasting Re
Figure 9 explains the valida
model through comparing the ou
from the model at certain days where the actual demand
profiles are known at those days.
Figure 10 indicates the mean square error that meas-
ures the deviation between the actual and forecasted de-
Figure 11 shows the curve fitting procedure that is
made through the ANN tool box during the forecasting
From the previous tables and curves the following
comments were noticed:
The load curve at peak day of year 2013 is main-
tained within the supply.
The load curves at peak days of years 2014 and
2015 are exceeding the supply capacity at certain times.
ting Results
ad curves of years
xceeds the s
The utility should put in the capacity expansion
plan starts from now-2012- so they could meet the de
mand requirements by the years 2014 and 2015.
The utility could apply one of the DSM alternatives
on the forecasted demand to achieve the best mh be-
The role of applying DSM LS algorithm in match-
ing the load with the supply.
tween the load and supply and this will be illustrated in
the following section.
6.2. DSM Load Shif
DSM load shifting is applied in the lo
2014 and 2015 as the load curve eupply
The importance of applying of integrating DSM and
load forecasting in the planning studies for utility com-
curve at the peak day in these years and the resultant
curves after applying DSM are as follows:
From the previous figures the following comments are
deduced as follows:
The peak demand is reduced.
The utility will postpone its investments to meet the
load requirements.
The Load Factor of the system is increased.
Figure 8. Historical Load curves.
Figure 9. Forecasting model validation.
Copyright © 2013 SciRes. EPE
Figure 10. ANN behavior Mean square error. Figure 11. Regression Curves from ANN tool box.
Figure 12. Peak day forecast at 2013, 2014 and 2015.
Figure 13. Load curves before and after shifting vs. day time in 2014.
Copyright © 2013 SciRes. EPE
Figure 14. Load curves before and after shifting vs. day time in 2015.
. Conclusions
In this paper a integrated scheme of DSM with load fo-
recasting is introduced over a practical case where the
following comments were noticed:
Impact of applying load forecasting and knowing
the peak demand.
The impact of DSM in matching load curve with the
supply curve.
The impact of DSM in reducing the peak demand of
the system.
The importance of applying of integrating DSM and
load forecasting in the planning studies for utility com-
The utility will postpone its investments in expand-
ing the generation capacities to meet the load.
[1] H. Attia, “Mathematical Formulation of the Demand Side
Management (DSM) Problem and its Optimal Solution,”
Proceedings of the 14th International Middle East Power
Systems Conference (MEPCON’10), Cairo University,
Egypt, December 19-21, 2010.
[2] Electric Power Research Institute, Demand-side Man-
agement: Utility Options for the Future, EPRI Reports,
CU. 3028.10.89, 2006.
[3] N. Kanoksing and W. Tayati, “Economics of Demand
Side Management and Hybrid Renewable Energy System
for a Remote Village Electrification in Northern Thailand,
AUPEC'07, Curtin University of Technology, Perth, Aus-
tralia, Dec 2007.
[4] R. Achnata, “Long Term Electric Load Forecasting using
Neural Networks and Support Vector Machines,” IJCST,
Vol. 3, No. 1, 2012.
[5] R. Holmukhe, “Artificial Neural Network based short
term load forecasting technique for Indian power system
energy management,” College of Engineering Bhanti dya-
peetb Diversity, Pune43, 2007.
[6] E. Feinberg, Applied Math & Statistics Stony Brook Uni-
versity NSF workshop, November 3-4, 2003.
[7] M. Peng, N. F. Hubele and G. G. Karady, “Advancement
in the Application of Neural Networks for Short-Term
Load Forecasting,” IEEE Transactions on Power System
RENCES Vol. 7, 1992, pp. 250-257.
[8] M. Abogaleela, M. Elsobki and M. ElMarsafawy, “A
Two Level Optimal DSM Load Shifting Formulation Us-
ing Genetics Algorithm, Case study: Residential load,
Power Africa 2012.IEEE PES, South Africa.
[9] Smart Meter “Iskra” Tool Owned by “TAQA Power”
Copyright © 2013 SciRes. EPE