Power System

Operation and Control

Session 5.
 Offline Reserve (scheduled reserve)
• Non-Spinning Reserve is off-line generation capacity that can be ramped to capacity and
synchronized to the grid within 10 minutes of a dispatch instruction by the ISO, and that is
capable of maintaining that output for at least two hours.
• Non-Spinning Reserve is needed to maintain system frequency stability during emergency
conditions.
• Non-spinning (unsynchronized) reserves act only after synchronisation.
• Units that can quickly respond to system events (such as the loss of a generator)
Quick-start diesel generating units, gas-turbine units,Hydroelectric power plant, pumped-storage
hydro-unit.
 Example
Suppose a power system consisted of two
isolated regions: a western region and an eastern
region. Five units have been committed to
supply 3090 MW.
The two regions are separated by transmission tie
lines that can together transfer a maximum of 550
MW in either direction.

What can we say about the allocation of spinning

reserve in this system?
 Minimum cost Unit capacity-Unit output

Loss of Unit1?
Loss of Unit4? The tie capacity of only 550 MW limits the transfer
 Thermal Unit Constraints
Minimum up time: Once the unit is running, it should not be turned off immediately.
Minimum down time: Once the unit is decommitted, there is a minimum time before it
can be recommitted.
Crew constraints: If a plant consists of two or more units, they cannot both be turned on at
the same time since there are not enough crew members to attend both units while starting
up.
Start-up cost: Before a thermal plant can feed electricity to the grid, it has to be started up,
i.e. ramped up at least to the minimum generation level. This usually comes at a cost
independent of how much output is produced
 Two approaches to treating a thermal unit
during its down period.
Start-up procedure of a turbine depends on its casing and rotor temperature.

Cold • Boiler has been off for

several days, zero pressure
Start on light-off.

Warm • Boiler is on hot stand-by

with some level of steam
Start pressure prior to light-off
 Time dependent start-up costs
Allows the unit’s boiler to cool down and
then heat back up to operating temperature
in time for a scheduled turn on.
Fuel Cost
Thermal time constant for the unit
Cold Start Cost when cooling= 𝐶𝐶𝐶𝐶 1 − 𝜀𝜀 −𝑡𝑡⁄𝛼𝛼 × 𝐹𝐹 + 𝐶𝐶𝑓𝑓

Cold-start cost (MBtu)

Time the unit was cooled
Fixed cost (includes crew expense, maintenance expenses )
Start-up cost when banking=𝐶𝐶𝑡𝑡 × 𝐹𝐹 × 𝑡𝑡 + 𝐶𝐶𝑓𝑓
Cost(Mbtu/h) of maintaining unit at operating temperature
 Other Constraints
• Must Run: A specific generating unit that has been designated by the system operator to be
on line or on the grid to insure the flow of electricity. A unit may be designated as must run for
operating reasons that may include system reliability, voltage control, or system stability.
• 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 unit commitment problem. (Some plants cannot be operated due to deficient fuel
supply.
• Emission Constraints: Air quality constraints in the form of limits on emissions from
fossil-fired plants
 Other Constraints
• Ramp rates: (The rate, expressed in
megawatts per minute, that
a generator changes its output.)To avoid
damaging the turbine, the electrical output
of a unit cannot change by more than a
certain amount over a period of time.
Ramp up-Ramp Down.
• Maximum generation capacity,
Minimum stable generation.
 Unit Commitment Solution Methods

• Solving the problem for M periods.

• There are N units to commit and dispatch.
• There are M load levels.
• N Operating limits for units.
• System constraints
 Complete enumeration
The total number of combinations we need to try each hour is:

𝐶𝐶 𝑁𝑁, 1 + 𝐶𝐶 𝑁𝑁, 2 +. . . +𝐶𝐶 𝑁𝑁, 𝑁𝑁 − 1 + 𝐶𝐶 𝑁𝑁, 𝑁𝑁 = 2𝑁𝑁 − 1

𝑁𝑁!
C(N,j) is the combination of N items taken j at a time 𝐶𝐶 𝑁𝑁, 𝑗𝑗 =
𝑁𝑁 − 𝑗𝑗 ! 𝑗𝑗!

For the total period of M intervals, the maximum number of possible combinations is 2𝑁𝑁 − 1 𝑀𝑀

For example, take a 24-h period (e.g., 24 one-hour intervals) and consider systems with 5, 10, 20, and
40 units
 Priority-List Method
Full-load average production cost of each unit:
Unit 1: Min=150MW
Max=600MW Unit1
Unit2 𝐻𝐻 𝑃𝑃max × 𝐹𝐹𝑢𝑢𝑢𝑢𝑢𝑢Cos𝑡𝑡
𝐻𝐻1 = 510 + 7.2𝑃𝑃1 + 0.00142𝑃𝑃12 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀⁄ℎ 𝐶𝐶𝑎𝑎𝑎𝑎𝑎𝑎 =
Unit3 𝑃𝑃max
Unit 2: Min=100MW
Max=400MW
𝐻𝐻2 = 310 + 7.85𝑃𝑃2 + 0.00194𝑃𝑃22 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀⁄ℎ

Unit 3: Min=50MW
Max=200MW
𝐻𝐻3 = 78 + 7.97𝑃𝑃3 + 0.00482𝑃𝑃32 𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀 ⁄ℎ

𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹Cos𝑡𝑡1 = 1.1 𝑅𝑅 ⁄𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀

𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹Cos𝑡𝑡2 = 1.0 𝑅𝑅⁄𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
𝐹𝐹𝐹𝐹𝐹𝐹𝐹𝐹Cos𝑡𝑡3 = 1.2 𝑅𝑅⁄𝑀𝑀𝑀𝑀𝑀𝑀𝑀𝑀
 Priority-List Method
Priority order for these units, based on the average production cost:

The commitment scheme would (ignoring min up/down time, start-up costs, etc.) simply use only the
following combinations:

When load is above 1000 MW

Between 1000 MW and 600MW
Below 600 MW
 Priority-List Method
• Check for enough capacity (sufficient generation to supply the load) +
spinning-reserve requirements.
• Check for the number of hours, H, before the unit will be needed again, i.e.
the load is dropping and will then go back up some hours later.
• If H is less than the minimum shut-down time for the unit, keep
commitment.
Which one is cost effective?

• Calculate two costs.

1. The sum of the hourly production costs for the next H hours with the unit
up.
2. Then recalculate the same sum for the unit down and add in the start-up
cost for either cooling the unit or banking it.
If there is sufficient savings from shutting down the unit, it
should be shut down, otherwise keep it on.
 Lagrange Relaxation Solution
If unit i is off-line during period t 𝑈𝑈𝑖𝑖𝑡𝑡 = 0
If unit i is on-line during period t 𝑈𝑈𝑖𝑖𝑡𝑡 = 1

Constraints: Non Linear term

𝑁𝑁𝑔𝑔𝑔𝑔𝑔𝑔

1. Loading constraints: 𝑡𝑡
𝑃𝑃𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 − � 𝑃𝑃𝑖𝑖𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡 = 0, 𝑓𝑓𝑓𝑓𝑓𝑓 → 𝑡𝑡 = 1, . . . . . , 𝑇𝑇
𝑖𝑖=1

2. Unit limits: 𝑈𝑈𝑖𝑖𝑡𝑡 𝑃𝑃𝑖𝑖min ≤ 𝑃𝑃𝑖𝑖𝑡𝑡 ≤ 𝑈𝑈𝑖𝑖𝑡𝑡 𝑃𝑃𝑖𝑖max , 𝑓𝑓𝑓𝑓𝑓𝑓 → 𝑡𝑡 = 1, . . . . . , T&𝑖𝑖 = 1. . . 𝑁𝑁

3. Other constraints: Unit minimum up- and down-time constraints, system air quality constraints in
the form of limits on emissions from fossil-fired plants, spinning reserve constraints.
 Lagrange Relaxation Solution
𝑇𝑇
𝑁𝑁

The objective function is: � � 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖𝑡𝑡 + 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆. Cos𝑡𝑡𝑖𝑖,𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡 = 𝐹𝐹 𝑃𝑃𝑖𝑖𝑡𝑡 , 𝑈𝑈𝑖𝑖𝑡𝑡
𝑖𝑖=1
𝑡𝑡=1

𝑁𝑁
Load balance: 𝑡𝑡
𝑃𝑃𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙 − � 𝑃𝑃𝑖𝑖𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡 = 0, 𝑓𝑓𝑓𝑓𝑓𝑓 → 𝑡𝑡 = 1,2, . . . 𝑇𝑇
𝑖𝑖=1

𝑇𝑇
𝑁𝑁
Lagrange function: 𝐿𝐿 𝑃𝑃, 𝑈𝑈, 𝜆𝜆 = 𝐹𝐹 𝑃𝑃𝑖𝑖𝑡𝑡 , 𝑈𝑈𝑖𝑖𝑡𝑡 + � 𝜆𝜆𝑡𝑡 𝑃𝑃𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙
𝑡𝑡
− � 𝑃𝑃𝑖𝑖𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡
𝑖𝑖=1
𝑇𝑇 𝑡𝑡=1 𝑇𝑇
𝑁𝑁 𝑁𝑁
Minimizing the Lagrange
𝐿𝐿 = � � 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖𝑡𝑡 + 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 − 𝑈𝑈𝑈𝑈Cos𝑡𝑡𝑖𝑖,𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡 + � 𝜆𝜆𝑡𝑡 𝑃𝑃𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙
𝑡𝑡
− � 𝑃𝑃𝑖𝑖𝑡𝑡 . 𝑈𝑈𝑖𝑖𝑡𝑡
function s.t constraints:
𝑖𝑖=1 𝑖𝑖=1
𝑡𝑡=1 t=1
 Lagrange Relaxation Solution
Rewritten as:
𝑇𝑇 𝑁𝑁 𝑇𝑇
𝑁𝑁 𝑁𝑁

𝐿𝐿 = � � 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖𝑡𝑡 + 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆 − 𝑈𝑈𝑈𝑈Cos𝑡𝑡𝑖𝑖,𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡 + � 𝜆𝜆𝑡𝑡 𝑃𝑃𝑙𝑙𝑙𝑙𝑙𝑙𝑙𝑙

𝑡𝑡
− � � 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖𝑡𝑡 . 𝑈𝑈𝑖𝑖𝑡𝑡
𝑖𝑖=1 𝑖𝑖=1
𝑡𝑡=1 𝑖𝑖=1 𝑡𝑡=1

Constant
Fixed

𝑁𝑁
𝑇𝑇

𝐿𝐿 = � � 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖𝑡𝑡 + 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆, Cos𝑡𝑡𝑖𝑖,𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡 − 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡

𝑡𝑡=1
𝑖𝑖=1
Separating the units from one another:
𝑇𝑇

� 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖𝑡𝑡 + 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆, Cos𝑡𝑡𝑖𝑖,𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡 − 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡

𝑡𝑡=1

It can be solved separately for each generating unit, without regard for what is happening on the other
generating units.

The minimum of the Lagrangian:

𝑁𝑁 𝑇𝑇

min𝑞𝑞 𝜆𝜆 = � min � 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖𝑡𝑡 + 𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆𝑆, Cos𝑡𝑡𝑖𝑖,𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡 − 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖𝑡𝑡 𝑈𝑈𝑖𝑖𝑡𝑡
𝑖𝑖=1 𝑡𝑡=1
Subject to:

𝑈𝑈𝑖𝑖𝑡𝑡 𝑃𝑃𝑖𝑖min ≤ 𝑃𝑃𝑖𝑖𝑡𝑡 ≤ 𝑈𝑈𝑖𝑖𝑡𝑡 𝑃𝑃𝑖𝑖max , 𝑓𝑓𝑓𝑓𝑓𝑓 → 𝑡𝑡 = 1. . . 𝑇𝑇

Considering the up- and down-time constraints.

 Solved as a dynamic programming problem in one variable.

Off

On

𝑆𝑆𝑖𝑖 is the start-up cost of unit 1

At the 𝑼𝑼𝒕𝒕𝒊𝒊 = 0 state, the value of the function to minimized equals zero.

At the state where 𝑼𝑼𝒕𝒕𝒊𝒊 = 0 , the function to be minimized is: 𝐦𝐦𝐦𝐦𝐦𝐦 𝑭𝑭𝒊𝒊 𝑷𝑷𝒊𝒊 − 𝝀𝝀𝒕𝒕 𝑷𝑷𝒕𝒕𝒊𝒊
Finding the minimum of this function by taking the first derivative

𝑑𝑑 𝑡𝑡 𝑡𝑡
𝑑𝑑 𝑡𝑡 𝑡𝑡
𝑡𝑡 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖 − 𝜆𝜆 𝑃𝑃𝑖𝑖 = 𝑡𝑡 𝐹𝐹 𝑃𝑃𝑖𝑖 − 𝜆𝜆 = 0
𝑑𝑑𝑃𝑃𝑖𝑖 𝑑𝑑𝑃𝑃𝑖𝑖

𝑑𝑑 𝑜𝑜𝑜𝑜𝑜𝑜
𝑡𝑡 𝐹𝐹 𝑃𝑃𝑖𝑖 = 𝜆𝜆𝑡𝑡 The solution to this equation
𝑑𝑑𝑃𝑃𝑖𝑖

There are three cases to be concerned with depending on the relation of 𝑷𝑷𝒊𝒊
𝒐𝒐𝒐𝒐𝒐𝒐
and the unit limits:

𝒐𝒐𝒐𝒐𝒐𝒐
1. If 𝑷𝑷𝒊𝒊 <𝑷𝑷𝒎𝒎𝒎𝒎𝒎𝒎
𝒊𝒊 then: min 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖 − 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖𝑡𝑡 = 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖min − 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖min
𝒐𝒐𝒐𝒐𝒐𝒐 𝑜𝑜𝑜𝑜𝑜𝑜 𝑜𝑜𝑜𝑜𝑜𝑜
2. If 𝑷𝑷𝒎𝒎𝒎𝒎𝒎𝒎
𝒊𝒊 < 𝑷𝑷𝒊𝒊 <𝑷𝑷𝒎𝒎𝒂𝒂𝒂𝒂
𝒊𝒊 then: min 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖 − 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖𝑡𝑡 = 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖 − 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖
𝒐𝒐𝒐𝒐𝒐𝒐
3. If 𝑷𝑷𝒊𝒊 <𝑷𝑷𝒎𝒎𝒂𝒂𝒂𝒂
𝒊𝒊 then: min 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖 − 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖𝑡𝑡 = 𝐹𝐹𝑖𝑖 𝑃𝑃𝑖𝑖max − 𝜆𝜆𝑡𝑡 𝑃𝑃𝑖𝑖max
 Adjusting λ

Adjustment factors are different:

Adjustment factor

𝑡𝑡 𝑡𝑡
𝑑𝑑 α = 0.002 when 𝑑𝑑λ
𝑑𝑑
𝑞𝑞(λ) is negative.
𝜆𝜆 = 𝜆𝜆 + 𝑞𝑞 𝜆𝜆 𝛼𝛼
𝑑𝑑𝑑𝑑
α = 0.01 when 𝑑𝑑λ
𝑑𝑑
𝑞𝑞(λ) is positive.

Minimizing the Lagrangian

 Lagrange Relaxation Solution
Lagrange relaxation method for the unit commitment problem is that it can relax
(ignore) each generator’s output dependency caused by the demand–supply balance
constraint so that a unit commitment of each generator is determined independently
by dynamic programming.

Lagrange relaxation method has been used for unit commitment scheduling of large-
scale power systems. This method provides approximate solutions to mixed integer
programming problems.

The unit commitment (UC) optimization is a large-scale, non-convex, and mixed-

integer linear optimization problem that is hard to solve

MILP: Mixed-integer linear Programming

Unit Commitment is the process of turning on (committing) resources to
meet load and other market requirements.

SECURITY CONSTRAINED UNIT COMMITMENT (SCUC) represents a powerful

scheduling tool used in electricity markets for day-ahead planning.
SCUC is a framework that combines two common algorithms in the power
industry: Unit Commitment (UC) and Economic Dispatch (ED) while
adding a new perspective to it - security.

System security involves practices designed to keep the system operating when components
fail.