Unit Commitment

Daniel Kirschen

Economic Dispatch
l
l
l

Given load
Given set of units on-line
How much should each unit generate to
meet this load at minimum cost?

Load

Unit Commitment
l

l
l

Given load profile

(e.g. values of the load for each hour of a day)
Given set of units available
When should each unit be started, stopped and how
much should it generate to meet the load at minimum
cost?

Load Profile

Typical
Summer
And
Winter
Demands

Why try to optimise?

l

Energy traded through the Electricity Pool of England and

Wales: ~ 7 billion per year
0.1% cost reduction through better scheduling:
~ 7 million

A Simple Example
l

Unit 1:
PMin = 250 MW, PMax = 600 MW
C1 = 510.0 + 7.9 P1 + 0.00172 P12 /h

Unit 2:
PMin = 200 MW, PMax = 400 MW
C2 = 310.0 + 7.85 P2 + 0.00194 P22 /h

Unit 3:
PMin = 150 MW, PMax = 500 MW
C3 = 78.0 + 9.56 P3 + 0.00694 P32 /h

What combination of units 1, 2 and 3 will produce 550

MW at minimum cost?
How much should each unit in that combination
generate?

Cost of the various combinations

1
Off
Off
Off
Off
On
On
On
On

2
Off
Off
On
On
Off
Off
On
On

3
Off
On
Off
On
Off
On
Off
On

Pmin
0
150
200
350
250
400
450
600

Pmax
0
500
400
900
600
1100
1000
1500

P1

P2
P3
Infeasible
Infeasible
Infeasible
0 400 150
550
0
0
400
0 150
295 255
0
Infeasible

Ctotal

5418
5389
5613
5471
5617

Observations on the example:

l

Far too few units committed:

Cant meet the demand
Not enough units committed:
Some units operate above optimum
Too many units committed:
Some units below optimum
Far too many units committed:
Minimum generation exceeds demand
No-load cost affects choice of optimal combination

Another Example
l

Optimal generation
schedule for a load
profile
Decompose the profile
into a set of period
Assume load is constant
over each period
For each time period,
which units should be
committed to generate at
minimum cost during that
period?

Load
1000

500

Time
0

12

18

Optimal combination for each hour

Load
1100
1000
900
800
700
600
500

Unit 1
On
On
On
On
On
On
On

Unit 2
On
On
On
On
On
Off
Off

Unit 3
On
Off
Off
Off
Off
Off
Off

24

Matching the combinations to the load

Load

Unit 3

Unit 2
Unit 1

12

18

24

Time

Issues
l

Must consider all constraints

n
n

l
l

Unit constraints
System constraints

Some constraints create a link between the periods

Starting up a generating unit costs money in addition to
the running cost considered in economic dispatch
Curse of dimensionality

Unit Constraints
l

Constraints that affect each unit

individually:
n
n
n
n
n
n

Maximum generating capacity

Minimum stable generation
Flexibility
Minimum up time
Minimum down time
Ramp rate

Flexible Plants
l
l

Power output can be adjusted (within limits)

Examples:
n
n
n
n
n

Coal-fired
Oil-fired
Open cycle gas turbines
Combined cycle gas turbines
Hydro plants with storage

Thermal units

Status and power output can be optimised

Inflexible Plants
l

Power output cannot be adjusted for technical or

commercial reasons
Examples:
n
n
n
n

Nuclear
Run-of-the-river hydro
Renewables (wind, solar,)
Combined heat and power (CHP, cogeneration)

Output treated as given when optimising

Notations
X i (t ) : Status of unit i at period t
X i (t ) = 1: Unit i is on during period t

X i (t ) = 0: Unit i is off during period t

Pi ( t )

: Power produced by unit i during period t

Unit Constraints
l

Minimum up time
n

Once a unit is running it may not be shut down

immediately:
If Xi (t) = 1 and tiup < tiup,min then X i (t + 1) = 1

Minimum down time

n

Once a unit is shut down, it may not be started

immediately

If Xi (t) = 0 and tidown < t idown,min then Xi (t + 1) = 0

Unit Constraints
l

Maximum ramp rates

n

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:

Pi (t + 1) Pi (t) Piup,max

Maximum ramp down rate constraint:

Pi (t) Pi (t + 1) Pidown,max

System Constraints
l

Constraints that affect more than one unit

n
n
n
n
n

Load/generation balance
Reserve generation capacity
Crew constraints
Emission constraints
Network constraints

System Constraints: Load/generation balance

Pi (t) = L(t)

iC(t)

C(t) = {i

X i (t) = 1}: Set of units committed at time t

System Constraint: Reserve Capacity

Unanticipated loss of a generating unit or an
interconnection causes unacceptable frequency drop if
not corrected
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

Pimax L(t) + R(t)

iC(t)

R(t): Reserve requirement at time t

How much reserve?

l
l

Protect the system against credible outages

Deterministic criteria:
n
n

Capacity of largest unit or interconnection

Percentage of peak load

Probabilistic criteria:
n

Takes into account the number and size of the

committed units as well as their outage rate

Types of Reserve
l

Spinning reserve
n

Primary

Secondary

quick response for a short time

slower response for a longer time
n

High frequency
ability to reduce output when frequency is high

Scheduled or off-line reserve

n

Other sources of reserve

n
n

Unit that can start quickly (e.g. gas turbines)

Pumped hydro plants
Demand reduction

Reserve must be spread around the network

Cost of Reserve
l

Reserve has a cost even when it is not called

n

More units scheduled than required

Units not operated at their maximum efficiency
Extra start up costs

n
n

Must build units capable of rapid response

Cost of reserve proportionally larger in small systems
Important driver for the creation of interconnections between systems

Crew Constraints
l

It may not be possible to start more than one generating

unit at a time in a power station because of the number of
people required to supervise the start-up
Less of a problem than it use to be thanks to automation

Emission Constraints
l

Amount of pollutants that generating units can emit may

be limited
Pollutants:
n

SO2, NOx

Various forms:
n
n
n

Limit on each plant at each hour

Limit on plant over a year
Limit on a group of plants over a period

Network Constraints
l

Transmission network may have an effect on the

commitment of units
n
n

Some units must run to provide voltage support

The output of some units may be limited because their
output would exceed the transmission capacity of the
network

Cheap generators
May be constrained off

More expensive generator

May be constrained on

Start-up Costs
l

l
l

Thermal units must be warmed up before they can be

brought on-line
Warming up a unit costs money
Start-up cost depends on time unit has been off
t OFF
i

SC i (t OFF
) = i + i (1 e i )
i

i + i
i
tiOFF

Start-up Costs
l
l

Need to balance start-up costs and running costs

Example:
n
n

Diesel generator: low start-up cost, high running cost

Coal plant: high start-up cost, low running cost

Issues:
n
n

How long should a unit run to recover its start-up cost?

Start-up one more large unit or a diesel generator to cover
the peak?
Shutdown one more unit at night or run several units partloaded?

Conclusion
l
l

Some constraints link periods together

Minimising the total cost (start-up + running) must be
done over the whole period of study
Generation scheduling or unit commitment is a more
general problem than economic dispatch
Economic dispatch is a sub-problem of generation
scheduling

Solving the Unit Commitment Problem

l

Decision variables:
n

Status of each unit at each period: Xi (t)

Xi (t) {0,1}!!!!i,t

Output of each unit at each period: Pi ( t )

Pi (t) 0, Pi min ; Pi max !!!i,t

l

Combination of integer and continuous variables

Optimisation with integer variables

l

Continuous variables
n
n

Can follow the gradients

Any value within the feasible set is OK

Discrete variables
n
n
n

There is no gradient
Can only take a finite number of values
Must try combinations of discrete values

How many combinations are there?

l

111
110

Examples
n 3 units: 8 possible states
N possible states
n N units: 2

101
100
011
010
001
000

How many solutions are there anyway?

l

T=

Optimisation over a time

horizon divided into intervals
A solution is a path linking one
combination at each interval
How many such path are
there?

How many solutions are there anyway?

l

Optimisation over a time

horizon divided into intervals
A solution is a path linking one
combination at each interval
How many such path are
there?
Answer:

(2 N )( 2 N ) K ( 2 N ) = (2 N )T
T=

The Curse of Dimensionality

l

Example: 5 units, 24 hours

(2 N )T = (2 5 ) 24 = 6.2
l

l
l

10 35 combinations

Processing 109 combinations/second, this would take 1.9

1019 years to solve
There are over 100 units in England and Wales...
Many of these combinations do not satisfy the constraints

How do you Beat the Curse?

Brute force approach wont work!
Need to be smart
Try only a small subset of all combinations
Cant guarantee optimality of the solution
Try to get as close as possible within a reasonable
amount of time

l
l
l
l

Main Solution Techniques

Priority list / heuristic approach

Dynamic programming
Lagrangian relaxation
Mixed Integer Programming

Characteristics of a good technique

l
l
l

n
n
n

Solution close to the optimum

Low computing time
Ability to model constraints