Energy System

Optimization
Modelling through
PLEXOS for SAARC
member states
Energy Exemplar
 Why Energy Exemplar?
1. Market Leader since 1999. 350+ clients in 62+ countries

2. No consulting. We’re focused on our platform.

3. 24x6 support worldwide

4. Strong investment in development

5. Social gateway to a global community

1. Network at User Group Meetings

2. Adopt Best Practices

3. Enjoy the knowledge transfer

3
 Energy Exemplar footprint
1,700 users at over 350 companies in 62 countries

Clients by Region:

1 9 18 124

Clients by Segment:
Utilities: 97
Consultants: 44
Power Producers: 39
Researchers: 28
Regulators: 21
TSOs: 14
ISOs: 13
Traders: 5
Energy Analysts: 4
International Institutes: 3
Asia Pacific Customers

5
Why optimization?

Evaluate
Manage
Uncertainty
multiple
scenarios

Optimized
forecasting Foresight
accuracy

Minimize
costs
Fuel savings

Is your existing approach optimal enough? 6

PLEXOS Integrated Platform
Why PLEXOS?
Best In Cutting-
Multi-task Transparent
Class edge

Mathematical Not a black box

Power
Optimization
Long
Medium
Short-term
Deterministic &
Gas Stochastic Customizable
Techniques

Water Robust Integrated Tools

8
 How PLEXOS works?

DATA SETS HIGHLY DETAILED PORTFOLIO DATA OUTPUT

Renewable Zonal Data • Heat rates • Solution review

Fuel Prices
Forecasts • Ramp rates, start profiles • Output reports
INPUT DATA SERIES

• Fuel blends and limits • Graphical interface

Energy Generation • Emissions constraints • Data export
Demand Availability Nodal Data • Trading positions
.zip
.xml
Water Fuel
Demand Contracts USER INTERFACE USER INTERFACE

.csv

Data read
Power system logic Engine Solver
Formulation & solution CPLEX, Xpress,
MIP and LP Gurobi, etc.
Solution interpretation
ENGINE

Reliability Evaluation
PSS®E AC OPF
PLEXOS Main Applications
Dispatch
Optimization

Reliability Hydro-thermal
Analysis Coordination

Generation
Energy and
and PLEXOS Reserves Co-
Transmission Functions optimization
Planning

Electricity and
Renewables
Gas Co-
Integration
optimization

Policy and
Strategy
Evaluation

10
 Integrated Energy Model
Post-Fact Hour Ahead to Next Month Next Month to Next 5 Years Next 5 Years to Next 30+ Years

Market Market Analysis: Understanding market landscape, trends, effects and forecasting

Long Term Integrated Resource Planning

Portfolio Medium Term Planning
Short Term Operations

Why single engine?

• Collaboration among institutions
• Co-optimize across use cases
• Performant for the most complex problems
• Role based UI
12
 Long-Term Planning

Minimum cost plan

Cost $
C(x): Capital Cost How Much
Which Capacity to
Tech to build/retire
build/retire

When Where
P(x): Production Cost To build/retire To build/retire

x = Assets
MIP Objective: Minimize NPV of capital & production
Total
C(x) P(x)
Cost
14
 Indicative Generation Capacity
Expansion Plan (IGCEP 2047)
Objectives:
Identify new generation requirements
by capacity, fuel technology and
commissioning dates on year-by-year
basis;
Satisfy the Loss of Load Probability
(LOLP) not more than 1% year to year;
Cater for the long-term load growth
forecast and reserve requirements; and
Provide a least cost optimal generation
expansion plan for development of
hydroelectric, thermal, nuclear and
renewable energy resources to meet the
expected load demand up to the year 2047
15
 Load Forecasting
Base Load Profile Peak and Energy Forecast
Energy (TWh) Peak Load (MW)
7000 Historical Load 2017
6500 90.0 12,000
6000

Peak Load (MW)

Energy (TWh)
5500 80.0
MW

5000 10,000
70.0
4500
4000 60.0
8,000
3500 50.0
40.0 6,000
2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028

Forecasted Load Profile

16
Renewable Energy Integration
PLEXOS® Stochastic Optimization considers multiple renewable profiles
to deliver a single optimal build decision.

Uncertainty doesn’t disappear

…it’s just accounted for properly
17
 GREENING THE GRID:
Pathways to Integrate 175 Gigawatts of
Renewable Energy into India’s Electric Grid

Higher RE penetration levels brings

Why Energy Exemplar: benefits of reduced fuel
consumption and lower
CO2 emissions.

A 175-GW RE target by 2022 that places

greater emphasis on wind over solar (60 GW
wind, 100 GW solar), achieves higher RE
capacity factors

18
 Medium Term Planning
LT Plan MT Schedule ST Schedule

Decompose inter-temporal
constraints that would otherwise
020 be too long for ST Schedule
500000
450000
Hydro Fuel
End Volume (1000 m3)

400000
350000 storage levels constraints
300000
250000 Inter-temporal
200000 Constraints
150000
100000 Generator
50000 technical Emission limits
0
limits

ST MT 20
System Security and Maintenance Planning
Projected Assessment of System Adequacy (PASA) is a simulation
phase that focuses on the balance of supply and demand in the
medium term. The functions of the PASA simulation phase are:

• To schedule maintenance events of any asset (transmission lines,

generators, gas pipelines).
EDNS

• Allocate outages events

Reliability
EENS LOLP
• Compute reliability statistics such as: Reporting

LOLE
21
Hydro Optimization
Multiple phase
Ensures optimal use of storage down to
simulation chronological level
(Long – Medium – Short)

Cascading Major and minor storages and junctions

networks Natural inflows and spillways and canals

Minimum releases for environment

Constraints Operational constraints and hydro generation
efficiency

Monte Carlo or Multistage stochastic optimisation for better

Stochastic modelling of storage release policies under
uncertainty
Optimisation
22
 Hydro Operation and Planning
2 interconnected systems connected through a 1,700 MVA Challenges:
2x500 kV line consisting in a system of 3,200 km extension

1) Transmission: High congestions in the

transmission lines due to large supply
of solar energy in the north to serve
high demand in the central zone

2) Cycling: Increase in startups

between 16 and 21 hours (Solar gen
decrease) and shutdowns between
Storage Volume Trajectory for 12 series of Natural Inflow
2500 6 and 12 hours (Solar gen increase).
2250
End Volume (1000 m3)

2000
1750 3) Ramping: High ramping
1500
1250 requirements with solar generation
1000
750
increase/decrease.
500
250
0
4) Hydro constraints: Highly complex
irrigation constraints.
Serie 1 Serie 2 Serie 3 Serie 4 Serie 5 Serie 6 23
Serie 7 Serie 8 Serie 9 Serie 10 Serie 11 Serie 12
 Short Term Formulation
Minimise Operational Costs + Start−up/Shutdown Costs

Subject to: Energy balance

Technical limits
System security requirements
When
How should each unit be
serve Load at started/stopped
minimum cost

How
How Much meet system
should each unit security
generate requirements
26
 Dispatch Optimization
Over 300+ Unit
Mixed Integer Commitment
Constraints Resolution down
Programing properties modelling to 1-second

Combined Cycle Gas Turbine (CCGT) Generation Profile of 1 week

700 Minimum Up Time Minimum Down Time Start Costs
600

500
Generation (MW)

400

300

200

100 Minimum Stable level

0
1 5 9 131721 1 5 9 131721 1 5 9 131721 1 5 9 131721 1 5 9 131721 1 5 9 131721 1 5 9 131721
1 2 3 4 5 6 7
Days
STEAM
CHILCA1_TV
GAS
CHILCA-TG1
GAS
CHILCA-TG2
GAS
CHILCA-TG3 27
TURBINE TURBINE 1 TURBINE 2 TURBINE 3
Reduce Fuel Costs
1. PLEXOS unit commitment algorithms work to minimize costs subject to constraints
2. CLP has many fuel constraints
1. Coal port and different mix ratios for coal plant in Hong Kong
2. Nuclear imports from China
3. Gas supply contracts with max daily quantity, take or pay etc.
3. CLP has different technologies like steam turbines, CCGTs and GTs
1. Over 300 UC properties in PLEXOS emulate life like dispatch
4. CLP has an emissions constraint for Hong Kong
1. The PLEXOS constraint class allows almost any constraint to be applied
2. The PLEXOS conditional class is like setting up “if” statements in excel
3. The PLEXOS variable class allows relationships to be assigned
5. These features together with Mixed Integer Programming emulate the real world.
6. Very significant savings are achieved

28
 Cascading System Optimization
Time: t Inflow6
Total Generation = 1255 MW 272.1 m3/s
300 m3/s
• Generators flow water from their
Inflow1
117.2 m3/s
Inflow2
Gen8
Head Storage to their Tail
34.84 m3/s
25.1 MW
~ ~ Gen10
2 MW Storage or ‘to the sea’.
Leaks Lake Lake
1 2
Gen9

272.9 m3/s
~
89.6 726.7
Hm3 Hm3 6 MW

Gen1 Gen2 Gen6

Irrigation Constraints
• A Waterway flows water from
~ 0 MW ~ 105 MW 0 MW
~ 2 m3/s

one Storage to another

Lake 3
120.1 Hm3
Inflow3 Inflow4
14.9 m3/s 126 m3/s
~ Gen11
550 MW
Inflow7
58.1 m3/s Irrigation Constraints
Gen3
~ • As the water runs downstream in
66.4 m3/s

42.9 MW 306.9 m3/s Leaks

Gen7 & Irrigation constraints
Lake

a cascade system, there is also a

34 MW
~
4
967.4

~
13.8 m3/s
Gen4 Hm3
53.5 MW
Gen12 loss of potential energy during
~ Gen15 Gen16

the trajectory which can be

298 MW 0 MW
Gen5
23.8 MW
~ Inflow5
113.8 m3/s
13.4 MW
~ ~
Gen13
64.2 MW ~ handle with PLEXOS
292.8 m3/s

~
Gen14
37 MW

Irrigation Constraints
192.7 m3/s
29
 Transmission Modelling
• Transmission operation
• Transmission Expansion Planning
• Security Constrained Economic
Dispatch (SCED)
• ‘N-x’ Contingency Analysis
• Congestion Costs
• Transmission aggregation and
Network reduction
• PSSE Power Flow Raw Data file
Importer Single Line Diagram (SLD) allow users to review
nodal connection

30
 Battery Energy Storage System
Solar Generation and Battery Net Generation Solar Moquegua
Solar farm
800
220
Battery
BESS Moquegua
1
600
220
Battery discharging Battery
BESS ILO
2 220
Generation [MW]

400

200

0
0 6 12 18 24 30 36 42 48
-200 Hours

-400
Battery charging

Built in best in BESS technical

class battery and economic
models representation

Represent all types of energy

storage technologies
32
 Ancillary Services Modelling
Static and/or Dynamic
Model any class of Define generators to
requirement for the
ancillary service provide reserve
service

• What price and hence cost was

imposed by reserve?

• Which generators and / or

interruptible load provide reserves?

• How reserve costs were allocated to

generators?

33
 Electricity and Gas Co-optimization
• Right-sizing for supply, storage, and transportation contracts
given gas and power demand
• Reliability requirements for delivering gas Gas Node 2
• Renewable generation impact studies to
serve both power and gas demand
• Storage injection and Gen_4

withdrawal target optimization Fuel_Gen_4

Gas Node 1

given both
power and gas demand Generator is
supplied with
PTG Fuel from the
Gas Node
 Integrated Strategic Optimization
• High & low renewable penetration
scenarios impact on gas demand
• Impacts of nuclear & coal retirements to
the gas demand
• Increasing carbon price impact on merit
order between gas and coal

• Modelling the injection of Biomethane

or Hydrogen into the gas system

• Power (excess wind power)to Gas

Production (Hydrogen) Scenarios
• Scenario Analysis- Electricity & Gas

35
Concluding Remarks

Better knowledge

Accurate foresight

A confident direction

Optimal operation

36
 Thank you for your kind attention
Do not hesitate to contact us if you have any question or want to know
more about our solutions.

Belen Leveroni
Business Development Executive
[email protected]

Dr. Dante Recalde

Lead Market Analyst
[email protected]

37
End