International Journal of Science and Research (IJSR)

ISSN (Online): 2319-7064

Index Copernicus Value (2013): 6.14 | Impact Factor (2015): 6.391

Unit Commitment by an Improved Priority List

Method
Karamjeet Kaur Bawa1, Amarjeet Kaur2
1
M.Tech student,Baba Banda Singh Bahadur Institute of Engineering and Technology,
Department of Electrical, Fatehgarh, Sahib, Punjab, India
3
Assistant Professor , Baba Banda Singh Bahadur Institute of Engineering and technology,
Fatehgarh Sahib,Punjab,India, Department of Electrical,

Abstract: Unit commitment problem helps in deciding which generation unit should be running in each period so as to satisfy a
predictably varying demand for electricity. Unit Commitment enables uninterruptible power to be delivered to consumers using the
principle of minimum operating cost. In this work, the unit commitment problem is solved using improved priority list method
approach. The generators are switched ON and OFF on a priority basis to minimize the total operating cost of the generating units. The
numerical results show that the improved priority list can commit the most economic unit first much better than other existing PL
methods.

Keywords: improved priority list method, unit commitment, priority list

1. Introduction developed. Several priority list methods have been developed

in the literature. The Unit Commitment (UC) is an important
The demand and supply of electricity need to be in constant research challenge and vital optimization task in the daily
balance. As demand for electricity has a typical weekly and operational planning of modern power systems due to its
seasonal pattern, power plants need to be carefully scheduled combinatorial nature. Because the total load of the power
to meet this fluctuating demand. This scheduling system varies throughout the day and reaches a different peak
optimization is known as the unit commitment (UC) problem value from one day to another, the electric utility has to
and has been widely discussed in the literature [1]. decide in advance which generators to start up and when to
Traditionally, the UC problem was solved centrally, connect them to the network and the sequence in which the
minimizing overall system cost. With the liberalization of operating units should be shut down and for how long. The
electricity markets worldwide, the aim is to operate the computational procedure for making such decisions is called
electricity generation systems with higher (economic) unit commitment. Unit commitment plans for the best set of
efficiency [2]. Focus has shifted to optimal economic units to be available to supply the predict forecast load of the
performance and profit maximization. On the one hand, the system over a future time period [9].
UC problem can be considered from a system’s perspective,
i.e., the so-called security-constrained UC. This type of UC is 2. Unit Commitment
similar to the traditional UC and is what an Independent
System Operator (ISO) currently deals with. Also towards 2.1 Problem formulation
policy making and planning, this UC is useful as a tool to
perform market simulations and assess the impact of specific In this section, we first formulate the UC problem. The
measures. On the other hand, from the viewpoint of a single objective of the UC problem is the minimization of the total
market player, a price-based UC problem can be considered, production costs over the scheduling horizon. Therefore, the
optimizing output towards maximum profit, based on objective function is expressed as the sum of fuel and start-up
electricity price forecasts [3, 4, 5 ].A wide range of solution costs of the generating units
techniques for the UC problem have been proposed and
developed over the years. Examples include priority listing The objective function of UC to be minimized is
(heuristics), Lagrangian relaxation, dynamic programming, Total cost = F ( P )+ Ci,t
i i,t
genetic algorithms, etc., together with hybrid methods Where:
combining several of these. Hence, the aim of this work is to F (P ) = ai+bi(P )+ci(P
i i,t i,t i,t
set up an adequate UC optimization tool that is able to cope The constraints include
with variable and low net demand profiles in an efficient (a) Limits of generation:
way. Such model is relevant for several market parties, all The produced power of each unit must obey the minimum
with their specific objectives. It is further usable for energy and maximum capacities, which can be defined as:
and climate policy evaluation and assessment. Also Pimin ≤ Pi,t ≤ Pimax
electricity generating companies can use this kind of model,
(b) Power-demand balance:
e.g., to provide operational solutions on relatively large sale,
The power produced by all committed units must meet the
or use the model in combination with other techniques such
system load demand, which is formulated as :
as MILP, to provide a starting solution. Towards this end, a
(i=1,N) Pi,t – PD = 0
new improved priority list (IPL) based method will be
Volume 5 Issue 5, May 2016
www.ijsr.net
Paper ID: NOV163787 1622
Licensed Under Creative Commons Attribution CC BY
 International Journal of Science and Research (IJSR)
ISSN (Online): 2319-7064
Index Copernicus Value (2013): 6.14 | Impact Factor (2015): 6.391
(c) Requirements of reserve: To maintain system reliability, 3. Priority List Method
adequate reserve requirement must be met at each period,
which is represented as: The goal of the PL is to commit the most economic unit first
Pmax ≥ Pi,t + PD to satisfy the constraints of the UC problem. The PL is often
(d) Limits of minimum up/down: used to solve UC neglecting the ramp rate constraints, thus,
T(i,on) ≤ X(i,on) only if the on/off status of a solution satisfies the spinning
T(i,off) ≤ X(i,off) reserve constraints and the minimum up/down time
Where: constraints, this method is feasible for UC without ramp rate
F (P ) = production cost function; constraints [10]
i i,t
Ci,t = startup cost;
ai,bi,ci = coefficients of the quadratic production cost of unit 4. Improved Priority List Method
i;
T(i,on) = minimum up time of unit i; The PL is easy to implement and has fast convergence rate.
T(i,off )= minimum down time of unit i; Nevertheless, it usually suffers from the highly heuristic
X(i,on) = duration during which unit i is continuously online; property and relatively poor quality solutions. The reason for
X(i,off) = duration during which unit i is continuously offline; the poor solutions is that the PL neglects the minimum
up/down time constraints when committing units to satisfy
2.2 Data specification the spinning reserve constraints. To make the initial solution
meet the minimum up/down time constraints, a series of
Table1: Generator system operator data heuristics are used to commit some new units order-commit
some on-line units, which is likely to violate the principle
committing the most economic unit first since some costly
units may be committed or some cheap units may be de-
committed to meet the minimum up/down time constraints.
As a result, the quality of the solutions obtained by PL is not
too high. To overcome these problems an advanced priority
list method is used.To improve the solution we now
investigate the solution without considering the MU and MD
constraints when turning on the generator based on the
priority order. The MU and MD constraints are considered in
calculation of startup cost to avoid the cold start cost when a
generator is turned off for more than the summation of MD
and cold start hour. The following are the core procedures in
MPL:
1) Turn on the generator in the order of priority
2) Check the MD from 1st to 24th hour in the order or
priority of the generators. If the generator is off between
two on states and the off duration is less than the MD, turn
on the generator at those hours. To compensate the extra
generated power, consider turning off generator from the
most expensive one if this does not violate the spinning
reserve, MU and MD constraints.
3) Check MU from 1st to 24th hour in the order of priority of
A system comprising of 10 generators is adopted as the test the generators. Turn on the generator for the next hour
bed in this work. The data specifications are given in Table1 until MU is satisfied. Once again, to compensate the extra
with the demand over 24-hour period available in Table2. 4) generated power, consider turning off generator from the
most expensive one if this does not violate the spinning
Table 2: Demand for 24 hours reserve, MU and MD constraints.

5. Methodology

a) Create a PL on the basis of Pmax value .if the value Pmax

of two units are same then check the heat rate of two units
and the unit with minimum value comes first in the priority
list.
b) Use the PL to commit units until the load demand plus the
spinning reserve are fulfilled at each period. Denote the
on/off status as an initial solution for UC without ramp rate
constraints. The total capacity of the committed generating
units must be bigger than or equal to the load and the
specified spinning reserve.
Volume 5 Issue 5, May 2016
www.ijsr.net
Paper ID: NOV163787 1623
Licensed Under Creative Commons Attribution CC BY
 International Journal of Science and Research (IJSR)
ISSN (Online): 2319-7064
Index Copernicus Value (2013): 6.14 | Impact Factor (2015): 6.391
c) Perform some heuristics to repair the initial solution to [3] Y. Tingfang and T. O. Ting, “Methodological Priority
meet the minimum up/down time constraints at period List for Unit Commitment Problem,” 2008 Int. Conf.
inserted into the load peak. Comput. Sci. Softw. Eng., no..2, pp. 176–179, 2008
d) Perform some to repair the solution obtained in Step 2 to [4] I. Damousis, “A solution to the unit commitment
meet minimum up/down time constraints of all periods. problem using integer-coded genetic algorithm,” IEEE
e) Calculate the production costs by economic load dispatch Trans. Power Syst., vol. 19, no. 2, pp.1165–1172, 2004.
(ELD) and startup costs by applying some heuristics. Add [5] T. Sum-im and W. Ongsakul, “Ant colony search
the both costs to get the total operation costs. algorithm for unit commitment,” in IEEE International
Conference on Industrial Technology, 2003, 2003, vol.
6. Results 1, pp. 72–77.
[6] Z.-L. Gaing, “Discrete particle swarm optimization
Table3: Solutions for improved priority list method algorithm for unit commitment,” in IEEE Power
Time in hours Production cost Startup cost Engineering Society General Meetin 2003, vol. 1, pp.
1 13683 0 418–424.
2 14554 0 [7] J. Ebrahimi, “Unit commitment problem solution using
3 16809 900 shuffled frog leaping algorithm,” IEEE Trans. Power
4 18589 0 Syst., vol. 26, no. 2, pp. 573– 581, 2011.
5 20020 560 [8] D. Srinivasan and J. Chazelas, “A priority listbased
6 22387 1100 evolutionary algorithm to solve large scale unit
7 23262 0 commitment problem,” in 2004.International Conference
8 24150 0 on Power System Technology, 2004 PowerCon2004.,
9 27251 860 2004, vol. 2, pp. 1746–1751.
10 30058 60
11 31916 60
12 33890 60
13 30058 0
14 27251 0
15 24150 0
16 21514 0
17 20642 0
18 22387 0
19 24150 0
20 30058 490
21 27251 0
22 22736 0
23 17684 0
24 15427 0
total cost 559887 4090

7. Conclusion
The improved priority list is simple and more efficient than
conventional priority list method. All the associated
constraints are met in the results.The Economic Dispatch
(ED) is solved using the lambda iteration method. The
simplicity of the improved priority list and fast calculation of
ED leads to a methodological and competent method in
comparison with conventional method. After calculation it is
concluded that the consideration of minimum up and down
constraints are necessary to minimize the overall cost.

References

[1] J. Wood and B. F. Wollenberg, Power Generation,

Operation, and Control, 2nd ed., vol. 37. New York:John
Wiley & Sons, Inc, 1996, p. 569.
[2] N. P. Padhy, “Unit commitment-a bibliographical
survey,” IEEE Trans.Power Syst., vol. 19, no. 2, pp.
1196–1205, May 2004.

Volume 5 Issue 5, May 2016

www.ijsr.net
Paper ID: NOV163787 1624
Licensed Under Creative Commons Attribution CC BY