Chapter 3

Unit Commitment
 Unit Commitment
Definition: The problem of finding the order in which the units
are to be brought in and the order in which the units are to be
shut down over a period of time, so that the total operating cost
involved on that day is minimum, is known as Unit Commitment
(UC) problem.

• When the load increases, the utility has to decide in advance

the sequence in which the generator units are to be brought
in.

• When the load decreases, the operating engineer need to

know in advance the sequence in which the generating units
are to be shut down.
 Unit commitment vs ED

OPF/ED Unit Commitment

Given load Given load profile

(e.g. values of the load for each
hour of a day)
Given set of units on-line Given set of units available

How much should each unit When should each unit be

generate to meet this load at started, stopped and how much
minimum cost? should it generate to meet the
load at minimum cost?
 Demonstration Example
Consider three thermal units with the following
necessary data.
• What combination of units 1, 2 and 3 will
produce 550 MW at minimum cost?
• How much should each unit in that
combination generate?

1 2 L
3
 Demontration Example
• Unit 1:
PMin = 150 MW, PMax = 600 MW
H1 = 510.0 + 7.9 P1 + 0.00142P12 MBtu/h
• Unit 2:
PMin = 100 MW, PMax = 400 MW
H2 = 310.0 + 7.85 P2 + 0.00194P22 MBtu/h
• Unit 3:
PMin = 50 MW, PMax = 200 MW
H3 = 78.0 + 7.97 P3 + 0.00482P32 MBtu/h

Hn – fuel input to unit n

 Example (Cont’d)
• Fuel cost:
Unit 1 = 1.1 $/MBtu
Unit 2 = 1.0 $/Mbtu
Unit 3 = 1.0 $/Mbtu

F1(P1)=H1(P1) x 1.1 = 561 + 7.92P1 + 0.001562P21 $/h

F2(P2)=H2(P2) x 1.0 = 310 + 7.85P2 + 0.00194P22 $/h
F3(P3) = H3(P3) x 1.0 = 78 + 7.97P3+ 0.00482P23 $/h
 Solution
• Try all combination of three units.
• For each feasible combination, units will be
dispatched using equal incremental cost rule studied
earlier. The results are presented in the Table below.
• Infeasible, if
𝑛

𝑃𝑚𝑎𝑥,𝑐𝑜𝑚𝑚𝑖𝑡𝑡𝑒𝑑 < 𝑃𝐷
𝑖
OR
𝑛

𝑃𝑚𝑖𝑛,𝑐𝑜𝑚𝑚𝑖𝑡𝑡𝑒𝑑 > 𝑃𝐷
𝑖
 Solution
1 2 3 Pmin Pmax P1 P2 P3 Ftotal
Off Off Off 0 0 Infeasible
Off Off On 50 200 Infeasible
Off On Off 100 400 Infeasible
Off On On 150 600 0 400 150 5418
On Off Off 150 600 550 0 0 5389
On Off On 200 800 500 0 50 5497
On On Off 150 1000 295 255 0 5471
On On On 300 1200 267 233 50 5617

Note 1: that the least expensive way of meeting the load is not with all the three
units running, or any combination involving two units. Rather it is economical to
run unit one alone.
Note 2: For each possible feasible combinations above calculate the cost using
incremental cost method discussed in the previous chapter and then find the
total cost.
 Continuation of the above example
Daily load curve to be met by a plant having three units is shown below.
Data pertaining to the three units are the same as in previous example.
Starting from the load of 1100 MW, taking steps of 100 MW find the shut-
down rule.

Load • Assume load is

constant over
each period
• Determine For
1000 each time
period, which
units should be
committed to
500 generate at
minimum cost
during that
period?
Time

0 6 12 18 24
Optimal combination for each hour

Load Unit 1 Unit 2 Unit 3

1100 On On On
1000 On On Off
900 On On Off
800 On On Off
700 On On Off
600 On Off Off
500 On Off Off
Matching the combinations to the load

Load

Unit 3

Unit 2

Unit 1

0 6 12 18 24
Time
 CONSTRAINTS ON UC PROBLEM
Based on the generation makeup and load-curve
characteristics different constraints can be placed on the unit
commitment problem.
Constraints:
– Unit constraints
– System constraints

• Unit Constraints (Constraints that affect each unit individually)

Minimum up time: once the unit is running, it should not
be turned off immediately.
Minimum down time: once the unit is decommited, there
is a minimum time before it can be recommitted
 Minimum up- and -down time
Notations:
u(i,t) : Status of unit i at period t
u(i,t) = 1: Unit i is on during period t
u(i,t) = 0 : Unit i is off during period t
x(i,t) : Power produced by unit i during period t

• Minimum up time
If u(i,t) = 1 and tiup < tiup,min then u(i,t +1) = 1
• Minimum down time

If u(i,t) = 0 and tidown < tidown,min then u(i,t +1) = 0

Ramp rates
• Maximum ramp rates
– To avoid damaging the turbine, the electrical output of a
unit cannot change by more than a certain amount over a
period of time:
Maximum ramp up rate constraint:
x ( i,t +1) - x ( i,t ) £ DPi up,max

Maximum ramp down rate constraint:

x(i,t) - x(i,t +1) £ DPi down,max

14
 System Constraints
• Constraints that affect more than one unit
– Load/generation balance
– Reserve generation capacity
– Environmental constraints
– Network constraints

15
Load/Generation Balance constraint
N

å u(i,t)x(i,t) = L(t)
i=1

N : Set of available units

Reserve Capacity Constraint
• Unanticipated loss of a generating unit or an interconnection
causes unacceptable frequency drop if not corrected rapidly
• Need to increase production from other units to keep
frequency drop within acceptable limits
• Rapid increase in production only possible if committed units
are not all operating at their maximum capacity
16
Reserve Capacity Constraint
N

å u(i,t )Pi
max
³ L(t ) + R(t )
i=1

R(t ): Reserve requirement at time t

The amount of reserve

• Protects the system against outages
• Capacity of largest unit
• Percentage of peak load

17
Types of Reserve
• Spinning reserve
– Primary
• Quick response for a short time
– Secondary
• Slower response for a longer time
• Tertiary reserve
– Replace primary and secondary reserve to protect
against another outage
– Provided by units that can start quickly (e.g. open
cycle gas turbines)

18
Types of Reserve
• Positive reserve
– Increase output when generation < load

• Negative reserve
– Decrease output when generation > load

• Other sources of reserve:

– Pumped hydro plants
– Demand reduction (e.g. voluntary load shedding)

19
 Environmental constraints
Scheduling of generating units may be affected by
environmental constraints

• Constraints on pollutants suchCO2 , SO2, NOx

– Various forms:
• Limit on each plant at each hour
• Limit on plant over a year
• Limit on a group of plants over a year
• Constraints on hydro generation
– Protection of wildlife
– Navigation, recreation
 Unit Commitment Solution Methods
Main solution techniques
• Priority list / heuristic approach
• Dynamic programming
• Lagrangian relaxation

Priority-List Method
In this method the full load average production cost
of each unit is calculated first. Using this, priority
list is prepared.
 Example
• Unit 1:
PMax = 600 MW
F1 = 510.0 + 7.9 P1 + 0.00172 P12 $/h
• Unit 2:
PMax = 400 MW
F2 = 310.0 + 7.85 P2 + 0.00194 P22 $/h
• Unit 3:
PMax = 500 MW
F3 = 78.0 + 9.56 P3 + 0.00694 P32 $/h

Obtain the priority list.

 Example

510+ 7.9∗600 + 0.00172∗6002

Full load avarage= = 9.782
600
production cost of unit 1

310+ 7.85∗400 + 0.00194∗4002

Full load avarage= = 9.401
400
production cost of unit 2

78+ 9.56∗500 + 0.00694∗5002

Full load avarage= = 13.186
500
production cost of unit 3
 Example
A strict priority order for these units, based on the average
production cost:

Unit $/h Max.MW The unit

2 9.401 400 Commitment
1 9.782 600 order
3 13.186 500

The shutdown scheme would simply use the following

combinations.
Combination Load PD
2+1+3 1000MW≤PD ∠1500MW
2+1 400MW≤ PD ∠ 1000MW
2 PD ∠ 400MW
 Priority based algorithm
• At each hour when the load is dropping, determine whether
dropping the next unit on the priority list will leave sufficient
generation to supply the load plus spinning reserve requirements. If
not, continue operating as is; if yes, go to next step.
• Determine the number of hours, H, before the unit will be needed
again assuming the load is increasing some hours later.
• If H is less than the minimum shut – down time for that unit, keep
the commitment as it is and go to last step; if not, go to next step.
• Calculate the two costs. The first is the sum of the hourly
production costs for the next H hours with the unit up. Then
recalculate the same sum for the unit down and add the start – up
cost. If there is sufficient saving from shutting down the unit, it
should be shut down; otherwise keep it on.
• Repeat the entire procedure for the next unit on the priority list. If
it is also dropped, go to the next unit and so forth.
Examples
• Given 3 units how many
111 possible combinations are
110 there?
101 – 3 units: 8 possible states
100 – N units: 2N possible states
011

010

001

000
How many solutions are there?

• Optimization over a
time horizon divided
into intervals
• A solution is a path
linking one
combination at each
interval
• How many such paths
are there?
T= 1 2 3 4 5 6

(23 −1)6 = 117649 combinations

Which Solution Techniques
• Characteristics of a good technique
– Solution close to the optimum
– Reasonable computing time
– Ability to model constraints

• Dynamic programming
• Lagrangian relaxation
Dynamic programming in unit commitment

If a system has n units, there would be 2n − 1

combinations which the dynamic programming
method enumerates (itemize) feasible schedule
alternatives and comparing them in terms of
operating costs.
Dynamic programming:
– Forward DP
• Start with present time and step by step decide until desired
time K
– Backward DP
• Start from the last time needed K and calculate to present time
 Dynamic programming
• Minimum cost in hour t with feasible state I is:

Where:
Ftc(t,I ): The total cost from initial state to hour t state I
Sc(t −1, L⇒ t,I): The transition cost from state (t −1, L) to state (t ,I)
{ L }: The set of feasible states at hour t − 1
F ( t , I ): The production cost for state ( t , I )

• The constraints
 Dynamic programming

• The constraints
 Dynamic programming

• We can combine the DP algorithm and priority list

method to discard some infeasible states as well as
high – cost states.
• Before we perform unit commitment using the
forward DP algorithm, we first order the units
according to the priority list.
• The first part of the units order is the must - up units,
the last part is the must – down units, and middle part
is the unit ranking based on the minimum average
production cost of the rest of units.
 Example
Using forward dynamic programming method solve the UC
problem of the four units given below. The load curve data for
8 hours is given in the following table.
Test system data
Ordering of the unit combinations or sates by
maximum net capacy
 UC results by priority list
Case 1 : Neglecting the constraints of unit minimum up/down time. Solve UC problem using
the priority list order.
Units are committed in order until the load is satisfied. The total cost for the interval is the
sum of the eight dispatch costs plus the transitional costs for starting any units. It can be known
from the average production cost in the table that the priority order for the four units is unit 3,
unit 2, unit 1, unit 4. All possible commitments start from state 12 since the load at first hour is
450 MW, and maximum net capacity from state 1 to state 11 is only 440 MW. In addition,
state 13 is discarded since it does not satisfy the order of the priority list. The UC results for the
priority ordered method are listed below.
 Example (Cont’d)
Case 2 : Neglecting the constraints of unit minimum up/down
time. Solve UC problem using dynamic programming.
First select the feasible states, using the priority list order. For first
4 hours, the feasible states have only 12, 14, and 15. For last 4
hours, the feasible states have 5, 12, 14, and 15. Thus the total
feasible states are {5, 12, 14, 15}, and the initial state is 12.
According to the recursive algorithm of the dynamic
programming, we can compute the minimum total cost
 Example (Cont’d)
• For the periods we select states
– For periods 1 to 4, the states will be [12,14 and 15]
– For periods 5 to 8, the states will be [5,12,14 and 15]
Now, compute the minimum total cost according to
the recursive algorithm

Starting state is {12}

For t=1, available states are {12, 14,15}
 Example (cont’d)
• For t=1, allowable states are {12, 14,15}

• For t=2,
 Assignment #3
Problem 5.2 (a and b) [1]

Submittion Date: April 18, 2020