Unit Commitment

W. D. Prasad
Department of Electrical Engineering

References:
[1] J. J. Grainger, W. D. Stevenson, “Power System Analysis”, McGraw Hill, ISBN: 978-0-07-058515-7
[2] S. Sivanagaraju, G. Sreenivasan, “Power System Operation & Control”
Introduction
• The total load of the power system varies throughout the day and reaches a
different peak value from one day to another.
• The electric utility has to decide in advance which generators to start up, when to
connect them to the network, sequence in which the operating units should be
shut down and for how long.
• The computational procedure for making such decisions is called the Unit
Commitment.
• When a unit is scheduled for connection to the system, it is said to be committed.

• Comparison with Economic Load Dispatch

Economic load dispatch economically distributes the actual system load as it
rises to the various units that are already on-line. However, the unit
commitment problem plans for the best set of units to be available to
supply the predicted or forecast load of the system over a future time
period.
Need for Unit Commitment
• The plant commitment and unit-ordering schedules extend the period of
optimization from a few minutes to several hours.

• Weekly patterns can be developed from daily schedules. Likewise, monthly,

seasonal, and annual schedules can be prepared by taking into consideration the
repetitive nature of the load demand and seasonal variations.

• A great deal of money can be saved by turning off the units when they are not
needed for the time. If the operation of the system is to be optimized, the UC
schedules are required for economically committing units in plant to service with
the time at which individual units should be taken out from or returned to
service.

• This problem is of importance for scheduling thermal units in a thermal plant; as

for other types of generation such as hydro, their aggregate costs (such as start-
up costs, operating fuel costs, and shut-down costs) are negligible so that their
on-off status is not important.
Constraints in Unit Commitment
• Spinning reserve
𝑡𝑜𝑡𝑎𝑙 𝑔𝑒𝑛𝑒𝑟𝑎𝑡𝑖𝑜𝑛 𝑜𝑓 𝑡𝑜𝑡𝑎𝑙 𝑙𝑜𝑎𝑑 +
𝑆𝑝𝑖𝑛𝑛𝑖𝑛𝑔 𝑅𝑒𝑠𝑒𝑟𝑣𝑒 = 𝑎𝑙𝑙 𝑠𝑦𝑛𝑐ℎ𝑟𝑜𝑛𝑖𝑧𝑒𝑑 𝑢𝑛𝑖𝑡𝑠 − 𝑡𝑜𝑡𝑎𝑙 𝑙𝑜𝑠𝑠𝑒𝑠
𝑎𝑡 𝑡ℎ𝑒 𝑖𝑛𝑠𝑡𝑎𝑛𝑡 𝑎𝑡 𝑡ℎ𝑎𝑡 𝑖𝑛𝑠𝑡𝑎𝑛𝑡
𝑛

𝑃𝐺,𝑆𝑅 = 𝑃𝐺𝑖 − 𝑃𝐷 + 𝑃𝐿
𝑖=1

• The spinning reserve must be maintained so that the failure of one or more units
does not cause too far a drop in system frequency. Simply, if one unit fails, there
must be an ample reserve on the other units to make up for the loss in a specified
time period.

• The reserves must be properly allocated among fast-responding units and slow-
responding units such that this allows the automatic generation control system to
restore frequency
Constraints in Unit Commitment
• Thermal unit constraints
A thermal unit can undergo only gradual temperature changes and this translates
into a time period (of some hours) required to bring the unit on the line. Due to
such limitations in the operation of a thermal plant, the following constraints are
to be considered.

a) Minimum up-time:
During the minimum up-time, once the unit is operating (up state), it should
not be turned off immediately.
b) Minimum down-time:
The minimum down-time is the minimum time during which the unit is in
‘down’ state, i.e., once the unit is de-committed, there is a minimum time
before it can be recommitted.

• Start-up Cost
The temperature and the pressure of the thermal unit must be moved
slowly, a certain amount of energy must be expended to bring the unit on-
line and is brought into the UC problem as a start-up cost.
Constraints in Unit Commitment
• Hydro unit constraints
• Operation of a system having both hydro and thermal plants is, complex as
hydro-plants have negligible operation costs, but are required to operate under
constraints of water available for hydro-generation in a given period of time.
• The problem of minimizing the operating cost of a hydro-thermal system can be
viewed as one of minimizing the fuel cost of thermal plants under the constraint
of water availability for hydro-generation over a given period of operation.

• Must run
• It is necessary to give a must-run reorganization to some units of the plant
during certain events of the year.

• Fuel constraints
• A system in which some units have limited fuel or else have constraints that
require them to burn a specified amount of fuel in a given time presents a most
challenging UC problem.
Problem Formulation
• Cost of operation 𝐹𝑖 of the 𝑖 𝑡ℎ unit can be given as;
𝐹𝑖 = 𝐶𝑖 𝑃𝐺𝑖
𝑃𝐺𝑖 = output of the 𝑖 𝑡ℎ unit
𝐶𝑖 = running cost of the 𝑖 𝑡ℎ unit (vary depending on the loading condition)

• Assume that the objective is to meet the load variation as indicated in the load
curve.
Problem Formulation
• Define the following variables.
𝐶𝑖𝑗𝑡 = variable cost coefficient of the 𝑖 𝑡ℎ unit when operating at the 𝑗𝑡ℎ load
for the sub-interval 𝑡
𝑃𝐺𝑖𝑗𝑡 = output of the 𝑖 𝑡ℎ unit at the 𝑗𝑡ℎ load for the sub-interval 𝑡
• Assume each unit is capable of operation at 𝑘 discrete levels. The running cost 𝐹𝑖𝑡
of the 𝑖 𝑡ℎ unit in the time interval 𝑡 is given as;
𝑘
𝐹𝑖𝑡 = 𝑗=1 𝐶𝑖𝑗𝑡 𝑃𝐺𝑖𝑗𝑡
• There are 𝑛 units available for the operation in the time interval 𝑡. Total running
cost of 𝑛 units during the time interval 𝑡 is given by;
𝑛 𝑘

𝐹𝑖𝑛𝑡 = 𝐶𝑖𝑗𝑡 𝑃𝐺𝑖𝑗𝑡

𝑖=1 𝑗=1
• For the entire time period of optimization, having 𝑇 sub-intervals of time, the
overall running cost of all the units become;
𝑇 𝑛 𝑘

𝐹𝑇 = 𝐶𝑖𝑗𝑡 𝑃𝐺𝑖𝑗𝑡
𝑡=1 𝑖=1 𝑗=1
Problem Formulation
• Start-up cost consideration
• Suppose that for a plant to be brought into service, an additional expenditure
𝐶𝑆𝑖 has to be incurred in addition to the running cost (i.e., startup cost of the
𝑖𝑡ℎ unit).
• The cost of starting ‘𝑥’ number of units during any sub-interval 𝑡,
𝑥

𝐹𝑆𝐶 = 𝐶𝑆𝑖 𝛿𝑖𝑡

𝑖=1
1 → 𝑖 𝑡ℎ 𝑢𝑛𝑖𝑡 𝑖𝑠 𝑠𝑡𝑎𝑟𝑡𝑒𝑑 𝑖𝑛 𝑠𝑢𝑏 − 𝑖𝑛𝑡𝑒𝑟𝑣𝑎𝑙 𝑡
𝛿𝑖𝑡 =
0 → 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

• Shut-down cost consideration

• If a plant is taken out of service during the scheduling period, it is necessary
to consider the shut-down cost.
• If ‘𝑦’ number of units are be to shut down during the sub-interval ‘𝑡’, the
shut-down cost is represented as;
𝑦

𝐹𝑆𝑑 = 𝐶𝑆𝑑 𝜎𝑖𝑡

𝑖=1
𝜎𝑖𝑡 = 1, when the 𝑖 𝑡ℎ unit is thrown out of service in sub-interval 𝑡.
Problem Formulation
• Total start-up cost over a period of 𝑇 sub-intervals
𝑇 𝑥

𝐹𝑆𝐶𝑇 = 𝐶𝑆𝑖 𝛿𝑖𝑡

𝑡=1 𝑖=1
• Total shut-down cost is given as;
𝑇 𝑦

𝐹𝑆𝑑𝑇 = 𝐶𝑆𝑑 𝜎𝑖𝑡

𝑡=1 𝑖=1

• Total expression for the cost function including the running cost, the start-up
cost, and the shut-down cost is written in the form:
𝑇 𝑛 𝑘 𝑇 𝑥 𝑇 𝑦

𝐹𝑇 = 𝐶𝑖𝑗𝑡 𝑃𝐺𝑖𝑗𝑡 + 𝐶𝑆𝑖 𝛿𝑖𝑡 + 𝐶𝑆𝑑 𝜎𝑖𝑡

𝑡=1 𝑖=1 𝑗=1 𝑡=1 𝑖=1 𝑡=1 𝑖=1

• For each sub-interval of time t, the number of generating units to be committed

to service, the generators to be shut down, and the quantized power loading
levels that minimize the total cost have to be determined.
Constraint Formulation
• Output of each generator must be within the minimum and maximum value of
capacity.
𝑃𝐺,𝑘,𝑚𝑖𝑛 ≤ 𝑃𝐺,𝑘 ≤ 𝑃𝐺,𝑘,𝑚𝑎𝑥

• Constraint on the total available capacity;

𝑛

𝑆𝑇𝐴𝐶 ≥ 𝑃𝐺𝑖,𝑚𝑎𝑥 𝛼𝑖𝑡 + 𝑆𝑟,𝑚𝑖𝑛

𝑖=1
𝑆𝑇𝐴𝐶 =total available capacity in any sub-interval 𝑡
𝑆𝑟,𝑚𝑖𝑛 = minimum running reserve capacity
1 → 𝑖 𝑡ℎ 𝑢𝑛𝑖𝑡 𝑖𝑠 𝑖𝑛 𝑜𝑝𝑒𝑟𝑎𝑡𝑖𝑜𝑛 𝑠𝑢𝑏 − 𝑖𝑛𝑡𝑒𝑟𝑣𝑎𝑙 𝑡
𝛼𝑖𝑡 =
0 → 𝑜𝑡ℎ𝑒𝑟𝑤𝑖𝑠𝑒

• Constraint on predicted load demand 𝑃𝐷 in sub-interval 𝑡

𝑛 𝑘

𝑃𝐺𝑖𝑗𝑡 ≥ 𝑃𝐷 + 𝑆𝑟,𝑚𝑖𝑛 + 𝑃𝐿
𝑖=1 𝑗=1
𝑆𝑟,𝑚𝑖𝑛 =minimum running reserve capacity
• The generator start-up and shut-down logical indicators, 𝛿𝑖𝑡 and 𝜎𝑖𝑡 should be
unity during the corresponding sub-intervals of operation.
Unit Commitment – Solution Methods
• Priority- List Schemes
• Dynamic Programming Method
• Lagrange’s Relaxation Method
• Evolutionary Optimization Techniques (GA, PSO, etc)
• Neural Networks, Fuzzy Logic, etc

• A power system with 𝐾 generating units must have at least one unit on-line to
supply the system load which is never zero over the load cycle. If each unit can be
considered either on (denoted by 1) or off (denoted by 0), there are 2 𝑘−1
candidate combinations.
• Let 𝐾 = 2;
Unit 𝒙𝟏 𝒙𝟐 𝒙𝟑
1 1 1 0
2 1 0 1

• Some combinations will be infeasible if the sum of all maximum MW for the units
committed is less than the load or if the sum of all minimum MW for the units
committed is greater than the load.
Example 1
• Consider a plant having 3 units. The cost characteristics, minimum and maximum
limits of power generation (MW) of each unit are indicated below:

𝑈𝑛𝑖𝑡 − 1: 𝐶1 = 0.002842𝑃𝐺1 2 + 8.46𝑃𝐺1 + 600 𝑅𝑠 ℎ𝑟 200 ≤ 𝑃𝐺1 ≤ 650

𝑈𝑛𝑖𝑡 − 2: 𝐶2 = 0.002936𝑃𝐺2 2 + 8.32𝑃𝐺2 + 420 𝑅𝑠 ℎ𝑟 150 ≤ 𝑃𝐺2 ≤ 450
𝑈𝑛𝑖𝑡 − 3: 𝐶3 = 0.006449𝑃𝐺3 2 + 9.884𝑃𝐺3 + 110 𝑅𝑠 ℎ𝑟 100 ≤ 𝑃𝐺3 ≤ 300

Incremental costs;
𝑑𝐶1 𝑅𝑠
𝑈𝑛𝑖𝑡 − 1: = 0.005684𝑃𝐺1 + 8.46 𝑀𝑊ℎ
𝑑𝑃𝐺1
𝑑𝐶2 𝑅𝑠
𝑈𝑛𝑖𝑡 − 2: = 0.005872𝑃𝐺2 + 8.32 𝑀𝑊ℎ
𝑑𝑃𝐺2
𝑑𝐶3 𝑅𝑠
𝑈𝑛𝑖𝑡 − 3: = 0.012898𝑃𝐺3 + 9.884 𝑀𝑊ℎ
𝑑𝑃𝐺3
Example 1
If all units are loaded at their minimum level, the incremental costs are given as;
𝑈𝑛𝑖𝑡 − 1, 𝑃𝐺1 = 200 𝑀𝑊, 𝜆1 = 9.5968 𝑅𝑠 𝑀𝑊ℎ
𝑈𝑛𝑖𝑡 − 2, 𝑃𝐺2 = 150 𝑀𝑊, 𝜆2 = 9.2 𝑅𝑠 𝑀𝑊ℎ
𝑈𝑛𝑖𝑡 − 3, 𝑃𝐺3 = 100 𝑀𝑊, 𝜆3 = 11.1738 𝑅𝑠 𝑀𝑊ℎ

Unit 1 Loading Unit 2 Loading Unit 3 Loading Total Load

(MW) (MW) (MW) (MW)

200 150 100 450

𝝀𝟏 = 𝟗. 𝟓𝟗𝟔𝟖 𝝀𝟐 = 𝟗. 𝟐 𝝀𝟑 = 𝟏𝟏. 𝟏𝟕𝟑𝟖
200 217.44 100 517.44
𝝀𝟏 = 𝟗. 𝟓𝟗𝟔𝟖 𝝀𝟐 = 𝟗. 𝟓𝟗𝟔𝟖 𝝀𝟑 = 𝟏𝟏. 𝟏𝟕𝟑𝟖

440.25 450 100 990.25

𝝀𝟏 = 𝟏𝟎. 𝟗𝟔𝟐𝟒 𝝀𝟐 = 𝟏𝟎. 𝟗𝟔𝟐𝟒 𝝀𝟑 = 𝟏𝟏. 𝟏𝟕𝟑𝟖

477.44 450 100 1063.44

𝝀𝟏 = 𝟏𝟏. 𝟏𝟕𝟑𝟖 𝝀𝟐 = 𝟏𝟎. 𝟗𝟔𝟐𝟒 𝝀𝟑 = 𝟏𝟏. 𝟏𝟕𝟑𝟖
Example 1
If all units are loaded at their minimum level, the incremental costs are given as;
𝑈𝑛𝑖𝑡 − 1, 𝑃𝐺1 = 200 𝑀𝑊, 𝜆1 = 9.5968 𝑅𝑠 𝑀𝑊ℎ
𝑈𝑛𝑖𝑡 − 2, 𝑃𝐺2 = 150 𝑀𝑊, 𝜆2 = 9.2 𝑅𝑠 𝑀𝑊ℎ
𝑈𝑛𝑖𝑡 − 3, 𝑃𝐺3 = 100 𝑀𝑊, 𝜆3 = 11.1738 𝑅𝑠 𝑀𝑊ℎ

Unit 1 Loading Unit 2 Loading Unit 3 Loading Total Load

(MW) (MW) (MW) (MW)

650 450 300 1400

𝝀𝟏 = 𝟏𝟐. 𝟏𝟓𝟒𝟔 𝝀𝟐 = 𝟏𝟎. 𝟗𝟔𝟐𝟒 𝝀𝟑 = 𝟏𝟑. 𝟕𝟓𝟑𝟒
Example 2
The system load in the figure below is to be supplied by combinations of the four
generating units given in the table. Treating Unit 1 and Unit 2 as must-run units,
determine the power supplied by the generators of each combination and the
corresponding production cost in economically loading the units when the system
load level is 1100 MW.

Load, (MW)
2000

1600

1200

800

400

Time, (h)
4 8 12 16 20 24
Example 2

Loading limits and fuel cost parameters

Generating Loading Limits Fuel Cost Parameters

Unit
Number Min Max 𝑎𝑖 𝑏𝑖 𝑐𝑖
(MW) (MW)
1 100 625 0.0080 8.0 500

2 100 625 0.0096 6.4 400

3 75 600 0.0100 7.9 600

4 75 500 0.0110 7.5 400

Answer – Example 2
Number of available generators – 4
Total number of possible combinations = 24 − 1 = 15

Unit 𝒙𝟏 𝒙𝟐 𝒙𝟑 𝒙𝟒 𝒙𝟓 𝒙𝟔 𝒙𝟕 𝒙𝟖 𝒙𝟗 𝒙𝟏𝟎 𝒙𝟏𝟏 𝒙𝟏𝟐 𝒙𝟏𝟑 𝒙𝟏𝟒 𝒙𝟏𝟓

1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0
2 1 1 1 0 0 0 1 0 1 1 1 1 0 0 0
3 1 1 0 1 1 0 0 0 1 1 0 0 1 1 0
4 1 0 1 1 0 1 0 0 1 0 1 0 1 0 1
Answer – Example 2
Number of available generators – 4
1
Average fuel cost of 𝑖 𝑡ℎ plant = 𝑓𝑖 = 𝑎𝑖 𝑃𝑔𝑖 2 + 𝑏𝑖 𝑃𝑔𝑖 + 𝑐𝑖 $
ℎ
2
Incremental fuel cost = 𝜆𝑖 = 𝑎𝑖 𝑃𝑔𝑖 + 𝑏𝑖 $ 𝑀𝑊ℎ
When two units are on economic dispatch:
𝜆 − 𝑏1 𝜆 − 𝑏2
𝜆 = 𝑎1 𝑃𝑔1 + 𝑏1 → 𝑃𝑔1 = ; 𝑃𝑔2 =
𝑎1 𝑎2
1 1 𝑏1 𝑏2
𝑃𝑔1 + 𝑃𝑔2 =𝜆 + − +
𝑎1 𝑎2 𝑎1 𝑎2
−1 −1
2 2 2
1 1 𝑏𝑖
𝜆= 𝑃𝑔1 + 𝑃𝑔2 +
𝑎𝑖 𝑎𝑖 𝑎𝑖
𝑖=1 𝑖=1 𝑖=1

General scenario with 𝑲 number of generators to meet the load:

𝝀 = 𝒂𝑻 𝑷𝒈𝑻 + 𝒃𝑻
−𝟏 −𝟏
𝑲 𝑲 𝑲 𝑲 𝑲
𝟏 𝟏 𝒃𝒊 𝒃𝒊
𝒂𝑻 = ; 𝑷𝒈𝑻 = 𝑷𝒈𝒊 ; 𝒃𝑻 = = 𝒂𝑻
𝒂𝒊 𝒂𝒊 𝒂𝒊 𝒂𝒊
𝒊=𝟏 𝒊=𝟏 𝒊=𝟏 𝒊=𝟏 𝒊=𝟏
Answer – Example 2
Combination – 𝑥7 1 1 0 0
Total load to = 1100 𝑀𝑊
−1 −1
1 1 1 1
𝑎𝑇 = + = + = 0.0044
𝑎1 𝑎2 0.008 0.0096
𝑏1 𝑏2 8 6.4
𝑏𝑇 = 𝑎 𝑇 + = 0.0044 + = 7.2727
𝑎1 𝑎2 0.008 0.0096
𝜆 = 𝑎 𝑇 𝑃𝑔𝑇 + 𝑏𝑇 = 0.0044 × 1100 + 7.3333 = 12.0727 $ 𝑀𝑊ℎ
𝜆 − 𝑏1 12.1733 − 8
𝑃𝑔1 = = = 509 𝑀𝑊
𝑎1 0.008
𝜆 − 𝑏2 12.1733 − 6.4
𝑃𝑔2 = = = 591 𝑀𝑊
𝑎2 0.0096
Fuel costs:
1
𝑓1 = 𝑎1 𝑃𝑔1 2 + 𝑏1 𝑃𝑔1 + 𝑐1 = 5609 $ ℎ
2
1
𝑓2 = 𝑎2 𝑃𝑔2 2 + 𝑏2 𝑃𝑔2 + 𝑐2 = 5858 $ ℎ
2
Fuel cost = 𝑓1 + 𝑓2 = $ 5609 + $5858 = 11467 $ ℎ
Total fuel cost for 4 hour duration= 45868 $
Answer – Example 2
Combination – 𝑥3 1 1 0 1 ; Total load to = 1100 𝑀𝑊
−1 −1
1 1 1 1 1 1
𝑎𝑇 = + + = + + = 0.0031
𝑎1 𝑎2 𝑎4 0.008 0.0096 0.011
𝑏1 𝑏2 𝑏4 8 6.4 7.5
𝑏𝑇 = 𝑎 𝑇 + + = 0.0031 + + = 7.3373
𝑎1 𝑎2 𝑎4 0.008 0.0096 0.011
𝜆 = 𝑎 𝑇 𝑃𝑔𝑇 + 𝑏𝑇 = 0.0031 × 1100 + 7.3373 = 10.774 $ 𝑀𝑊ℎ
𝜆 − 𝑏1 10.774 − 8
𝑃𝑔1 = = = 347 𝑀𝑊
𝑎1 0.008
𝜆 − 𝑏2 10.774 − 6.4
𝑃𝑔2 = = = 456 𝑀𝑊
𝑎2 0.0096
𝜆 − 𝑏4 10.774 − 7.5
𝑃𝑔4 = = = 298 𝑀𝑊
𝑎4 0.011
1
Fuel costs:𝑓1 = 𝑎1 𝑃𝑔1 2 + 𝑏1 𝑃𝑔1 + 𝑐1 = 3758 $ ℎ
2
1
𝑓2 = 𝑎2 𝑃𝑔2 2 + 𝑏2 𝑃𝑔2 + 𝑐2 = 4316.5 $ ℎ
2
1
𝑓4 = 𝑎4 𝑃𝑔4 2 + 𝑏4 𝑃𝑔4 + 𝑐4 = 3123 $ ℎ
2
Total Fuel cost = 4 × 𝑓1 + 𝑓2 + 𝑓4 = $ 44790
Answer – Example 2
Combination – 𝑥2 1 1 1 0 ; Total load to = 1100 𝑀𝑊

Total Fuel cost = $ 45848

Combination – 𝑥1 1 1 1 1 ; Total load to = 1100 𝑀𝑊

Total Fuel cost = $ 45848