october 2012 - ieee · pdf filefor transmission security limits it is assumed that...
TRANSCRIPT
1
October 2012
The Electric Power System includes three apects
2
Generation Transmission Distribution
Introduction
� Increasing competition and complexity in power generating systems� Necessity for high service reliability� Having low production costs
Triggers interests in automatic maintenance scheduling techniques
3
� Automatic maintenance scheduling can be done using� integer programming� decomposition methods� dynamic programming� simulated annealing method� probabilistic approach� artificial intelligence method
� This work represents a model for long-term preventive generationmaintenance scheduling problem using General Algebraic Modeling System.
� Long-term maintenance scheduling (LTS) considers the schedule ofgenerating units on a horizon of one or two years.
� LTS tackles fuel allocation, emission, budgeting, production, andmaintenance costing.
Introduction
4
The long-term Maintenance Scheduling problem is determining the periods for whichgenerating units should be taken off the line in order to minimize the total operation
cost . It can be formulated as follow:
∑∑ ∑∑ ∑ it
t i
it
t i
ittit gc)x(1 γCMint
ttFρ ×++
Maintenance Cost
Energy Production
Cost
Cost of Energy Purchased from
Outside
Cit: Maintenance cost of unit i at time t
γt: Weekly penalty factor
xit: Unit maintenance status, 0 if unit is off for maintenance, otherwise 1
cit: Generation cost of unit i at time t
git: Power generation for unit i at time t
Ρt: Cost of energy purchased from outside at week t
Ft: Expected Energy Purchased from outside
Problem Definition
5
Maintenance Constraints:
�Penalty factor: depicts importance of loading points based on amount of consumptions. ISO
could employ penalty factor to patronize unit not to have maintenance in peak loads.
�Maintenance window: The generation maintenance may not be scheduled before their
earliest period, or after latest period allowed for maintenance.
�Maintenance Duration: The maintenance of the generator i lasts a given amount of time.
�Maintenance Limitation: A maximum number of maintenance is imposed in the period t.
�Non-Stop Maintenance: The maintenance of a generator is carried out in consecutive periods.
�Exclusion Constraint: Generators i and j can not be in maintenance at the same time.
�One-Time Maintenance: each unit has an outage for maintenance only once along the time
horizon considered.
�Manpower Availability: in each maintenance area, there is limited available manpower for
maintenance.
Constraints
6
Network Constraints:
�Unit Capacity Limit: Each generator is designed to work between a
minimum and maximum power capacity (MW)
�Transmission Flow Limit: The power flows on transmission lines are
constrained by line capacity.
�Spinning Reserve: The reserve is the power provided if a generator fails. It
is a safety margin usually given as a demand proportion.
Constraints
7
�Due to discrete nature of model, mixed integer programming (MIP) is applied tosolve the problem.
�For this purpose General Algebraic Modeling System (GAMS) is the utilized forsolving optimization problem.
Solution Methodology
8
Generators’ data
Unit 10, 11 12, 13 14 15, 16 6, 7
Capacity (MW) 2×76 2×76 1×100 2×100 2×20
Bus 1 2 7 7 1
The proposed method is applied to the 24 bus IEEE-RTS.
24 Bus IEEE-RTS
Validation
� 32 generating units� 20 consumer� 24 buses and 38 transmission lines� A three month study period of summerweeks, weeks 18-29, is taken into account.� The maintenance area coverage is frombuses 1 through 10.� During three months, manpowerconstraint is up to two groups for unitmaintenance.� The maintenance windows are twoweeks for all generations.
9
Case 1: system reserve in each week is limited to the minimum of 1% and 6%, of the total weekly load.
$66,409,800
$66,410,000
$66,410,200
$66,410,400
$66,410,600
$66,410,800
$66,411,000
Study with 6% reserve Study with 1% reserve
Comparison of total operation & maintenance cost for case 1
Weekly peak load in percent of annual peak
Results: Case 1
Penalty Factor for generator unit maintenance cost
10
Unit\
Week
T18 T19 T20 T21 T22 T23 T24 T25 T26 T27 T28 T29
6
7
10
11
12
13
14
15
16
Unit\
Week
T18 T19 T20 T21 T22 T23 T24 T25 T26 T27 T28 T29
6
7
10
11
12
13
14
15
16
Maintenance scheduling of generating unit for System Reserve of 1%
Results: Case 1
Maintenance scheduling of generating unit for System Reserve of 6%
11
Case 2: studies the effect of transmission security, fuel limit and energy purchased from outside on maintenance scheduling problem. Here, system reserve in each week is limited to the minimum of 1% of the total weekly load.
For transmission security limits it is assumed that transmissions capacity between twobuses (15 to 21) are reduced to quarter.
For fuel limits are imposed on maintenance scheduling problem. The cost of per MWh ofenergy purchased from outside recourses is 49 $/MWh.
Results: Case 2
Study on fuel limits and
energy purchased from outside with 1% reserve
Study on the effect of
transmission security with 1% reserve
Study with 1% reserve
$79,815,950
$66,485,020$66,410,190
Comparison of total operation & maintenance cost for case 2
12
Unit\
Week
T18 T19 T20 T21 T22 T23 T24 T25 T26 T27 T28 T29
6
7
10
11
12
13
14
15
16
Unit\
Week
T18 T19 T20 T21 T22 T23 T24 T25 T26 T27 T28 T29
6
7
10
11
12
13
14
15
16
Maintenance scheduling of generating unit with limits on transmission line capacity 15-21 (case 2)
Maintenance scheduling of generating unit with limits fuel and energy purchased from outside (case 2)
Results: Case 2
13
This paper presents generation maintenance scheduling considering:
�Reserve of the system as reliability indices
�Network constraints
�Maintenance constraints
�Fuel constraints
�Energy purchased from outside
�Transmission capacity
�Penalty factor
In a global maintenance-scheduling problem, we propose to consider:
�Transmission maintenance scheduling
�Air pollution constraints
Conclusion and Future Work
