PLEXOS applications

PLEXOS
A P P L I C A T I O N S
®

Contents
Market Analysis
Page 3
̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣
Portfolio Optimization
Page 7
̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣
Reliability Analysis
Page 11
̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣
Capacity Expansion Planning
Page 15
̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣
Hydro Optimization
Page 19
̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣
Transmission Analysis
Page 24
̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣
Gas System Planning
Page 28
̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣
Renewable Generation Integration
Page 32
̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣̣

Market Analysis
In the past, the US electricity markets were characterized by vertically integrated and monopolistic
electric utilities. These utility companies owned the generation, transmission and distribution networks
and electric rates were reviewed and approved by public utility commissions or public boards.
However, in the past couple of decades, these vertically integrated structures have been replaced by
regional wholesale markets.
A number of regions, including the Northeast (NYISO, ISO-New England), Mid-Atlantic (PJM
Interconnection), much of the Midwest (Midwest ISO and Southwest Power Pool), Texas (ERCOT) and
California (CAISO) - organize their markets under an independent system operator (ISO) - sometimes
also referred to as a regional transmission organization (RTO). By adopting this ISO/RTO structure, these
regions have moved to expand competition in electricity from the vertically integrated to a competitive
markets for energy, capacity and ancillary products.
In fact, two-thirds of the electricity consumed in the U.S. are provided an ISO/RTO. Other regions -
including the Southeast (FRCC, SERC, TVA), Inter-Mountain West and Northwest - chose to retain the
traditional regulatory model. Under this regime, vertically-integrated utilities retain functional control
over the transmission system and therefore choose which generator is dispatched to serve load. Some
argue that such a model has led to preferential treatment by these utilities for their own generation
rather than more affordable and environmentally responsible generation available from competitive
suppliers and marketers.
U.S. natural gas markets are also highly competitive. Natural gas market prices are determined
competitively on spot and futures markets reflecting current and expected supply and demand
conditions. The market price is determined through the actions of thousands of well-informed buyers
and sellers. U.S. natural gas markets comprises of Midwest, Northeast, Gulf, Southeast and Western gas
regions.
ISO Market Structure and Products
The energy markets operated by ISO’s US typically include a two settlement market: day ahead market,
or financial commitment, and a real time or spot market. The day-ahead allows generators to
reasonably know whether they will be committed at least 12 hours before the start of the electric day.
Day ahead commitments are similar to a binding forward financial contract, where generators are
financially responsible for their dispatch the day prior. If they are unable to meet the dispatch in the real
time, or spot market, generators would have to purchase the difference between their day-ahead and
real time positions. The day-ahead market is normally an hourly market while the real time or spot
market is normally a 5 minute market (prices are settled and posted every 5 minutes).
Generators bid into both the day-ahead and real time markets using a bid curve to approximate their
marginal costs (typically their fuel and some notional operating costs). They are also allowed to provide
start-up costs for their commitment as well as minimum up and down times. All of these factors are
included in the algorithms established by the ISO to calculate the hourly price of electricity. The price
is set by the most expensive unit in the generator bid stack or the marginal unit.
energyexemplar.com Page 3

In addition to the day-ahead and real time energy markets, which indicate the price of energy required
to serve electric load over time, ISO’s have also devised their markets to provide locational price signals
as well. In this case, ISO’s operate a nodal electric market where the price of energy is determined by
the availability of transmission to deliver electricity across the network. In areas where there is
insufficient transmission to deliver all of a region’s demand, prices tend to increase to reflect the
increased costs to supply load. This is also known as congestion pricing. In the end, prices for
electricity are published both across time as well as across the network in which the ISO operates.
These prices are fully transparent and provide all market sectors the appropriate signals about the
available supply and demand of electricity.
Generators in the ISO operated markets submit bids which are typically capped at a certain or
prescribed level (roughly $1000 to $1200/MWh). As generators bids are capped and therefore they are
less likely to recoup their fixed costs in the energy market, ISO also provide generators with a capacity
market. These capacity markets are conducted in a bid auction for the right to supply generation
capacity. Generators are paid to be able to meet certain reliability criteria. These markets should also
provide sufficient incentive for new generation to enter the market, particularly when capacity and
energy prices increase over time.
Generators also bid into different ancillary services markets. These could include regulation, spinning
reserve, non-spinning or non-synch reserve and replacement reserves.
Lastly, the ISO also operate and manage the high voltage transmission network. Financial transmission
rights (FTR) / Congestion Revenue Rights (CRR) market.
PLEXOS® Integrated Energy Model
The markets served by ISO’s have grown increasing complex. With this complexity requires new
software tools for market participants of the deregulated markets. PLEXOS provides of software solution
to match the complexities.
PLEXOS offers multiple horizon simulations, including sub-hourly, to be able to model both the
day-ahead and real time spot markets. PLEXOS supports multiple spatial analyses, from a full nodal
network model to a zonal or regional model. As such, it is capable of calculating the system marginal
price, transmission congestion costs and losses, and other market metrics published by the ISO’s.
PLEXOS offers the same algorithms that the ISO use to dispatch their markets and is often used by the
ISO themselves for internal and external market studies.
Users of PLEXOS can create a multi-band bid and offer curves for both generation and demand. They
can create unit commitment constraints to match actual market decisions including minimum generation,
start costs and minimum up and down times. PLEXOS allows generation bid caps as well as the regional
price of value of loss load (VoLL). All of PLEXOS’s feature rich properties allow users to capture the
complexities of the US deregulated electricity markets in a single software platform.

Additional properties available in PLEXOS to address the deregulated electricity markets as operated
by the ISO’s are outlined in the table below.
Planning Objectives PLEXOS Capability
Renewables Integration and System
Flexibility Requirement Assessments
Least Cost Resource Change within
and Across Regions
Minimizing production costs and
consumer costs to electricity and
natural gas rate payers
Sizing Natural Gas Network
Components and Natural Gas
Storage
Environmental Policies
Integrated Reliability Evaluation
• Sub-Hourly Co-Optimization of Ancillary Services with Energy
Market and Transmission Power Flows
• Stochastic Optimization and Stochastic Renewables Models
• PHEV, EE, DR, SG, Energy Storage Models
• Co-Optimization of Generation and Transmission Expansion
• Generation Retirements and Environmental Retrofit Models
• Reliability Evaluation
• Co-Optimization of Production cost of Electrical and Natural
Sectors
• Electrical Network Contingencies and Natural Gas Network
Contingencies
• Co-Optimization of Natural Gas Network Expansion along with
Electricity Sector Expansion
• Electrical Network Contingencies and Natural Gas Network
Contingencies
• Co-Optimization of Annual and Mid-Term constraints
• Integrated Reliability Evaluation to Ensure LOLE and other Metrics
Maintained with Co-Optimization of Electric and Gas Sector
Expansion or True Monte Carlo
US PLEXOS® Integrated Datasets for Power and Natural
Gas Sector Challenges
In the past few years there has been a convergence in environmental policy and public policy, in
combination with recent shale gas developments which have combined to increased reliance of the
power sector on natural gas generation. As a result, gas and electric coordination has emerged as a
complex topic for regulators and market operators with concerns of potential gas constraints that impact
electric system operation and reliability. Many studies have been initiated for gas electric coordination
in the major interconnects (Eastern Interconnect, ERCOT and WECC) of the US. Furthermore, integration
of renewables have driven Integrated Resource Plan (IRP) managers and ISO’s to consider sub-hourly
ancillary co-optimization analysis in production cost models for major areas of the US. Thus the study
process and the complexity of issues is driving the demand for integrated datasets and integrated
models. In other words, one software package that can handle all the complexities of planning and
operations and market analysis in the short, medium and long term.
Energy Exemplar has developed integrated datasets serving North America, combining both a nodal
electric model as well as the natural gas network. We have three separate electric interconnection
models, the Eastern Interconnect (EI), Texas (ERCOT) and the Western Interconnect (WECC).
On top of these three models is a gas model serving North America. The gas network includes the full
scope of gas production and transportation system, from well head or production field to natural gas
pipelines, storage and end user demands.

The intersection of these two models is the natural gas fired electric generator. With the combination of
the gas electric dataset into a single model by PLEXOS, users can evaluate the combination of gas
constraints on the electric system, particularly in the winter when natural gas demand peaks, as well as
the interaction of price between the electric and gas models.
Planning Objectives PLEXOS Capability
Production
Cost
Annual Environmental Constraints
Mid Term
Optimization
Capacity
Expansion
Planning
Reliability
Evaluation
Generation Characteristics
Electrical Network Data
Natural Gas Network Data
Demand and Energy Shapes
Fuel Prices
Hurdle and Wheeling Rates
Constraints
Others
Energy Constraints
Mid Term Optimizations
Others
Integrated Planning
Database in PLEXOS
Generator Costs
Transmission Line Costs
Pipeline Costs
Gas Storage Costs
Energy Efficency Cost Curve
Demand Shapes
Others
Single Database
Facilitates Multiple
Types of Studies
FOR Transmission
FOR Generation
FOR Gas Pipeline
Renewable Variability
and Uncertainties
Demand Uncertainty
Others
Testimonial
“I've only been using PLEXOS for a couple of months, but I'm very excited about the flexibility and ease
with which I can test a wide variety of different scenarios. By only supplying key inputs I get surprisingly
accurate results, which I can confidently stand behind. So far PLEXOS has had more to offer than I've
needed, and each time when I run into a dead end I get quick responses with from a diligent technical
support.”
Jacobus Swart
NamPower
Africa

Portfolio Optimization
Portfolio Optimization includes the following four activities.
1. Operation. On a daily basis, the operators in the control center (or on the trading floor) solve unit
commitment and economic dispatch problems to minimum system costs while maximizing the profit
from trading with neighbouring utilities or markets.
2. Generating Asset Evaluation. Examples of generating assets include generators, energy contracts,
demand response programs, energy storage units, etc. Performing the Generating Asset
Evaluation involves quantifying the value of generating assets. Simulations both with and without
the generating asset are performed for the technical life time of the generating asset. Then, the
generating asset value is calculated as the production cost difference between the solutions with
and without the generating asset. Generating Asset Evaluation can also be used to examine the
difference between the generating asset operation and the energy and Ancillary Service markets.
3. Budgeting. Portfolio simulation can be used to predict the production cost, fuel consumption,
traded energy, emission production, etc. Solutions can then be used to support decisions on rate
design, fuel procurements, emission allowance allocation, etc.
4. Long-term Capacity Expansion. Long-term Capacity Expansion planning can be done using
simulations that determine optimal generation or transmission facility investment (or retirement) to
meet projected load growth while still complying with reliability criteria, emission limits,
renewable energy portfolio standard (RPS), etc.
Portfolio Optimization by PLEXOS®
PLEXOS solves Portfolio Optimization problem using the Security Constrained Unit Commitment (SCUC)
and Economic Dispatch (ED) algorithm. This algorithm makes use of Mixed Integer Programming (MIP)
to minimize a cost function subject to all operational constraints. The cost function may include the
generating cost, generator startup and shutdown costs, sales contract revenue, purchase contract cost,
energy or AS market sales revenue and purchase cost, transmission wheeling charges, etc. The
constraints can include things such as energy balance (i.e., at any moment, the generation plus
purchase are equal to the load plus sales), AS requirements, generating asset chronology (i.e., min up
/ down time, ramp rate, energy resources (i.e., energy limits, fuel limits, emission limits, water limits),
transmission limits, etc. For example, Portfolio Optimization can be formulated as follows:
Minimize Portfolio Production Cost = generator fuel and VOM cost + generator start cost
+ contract purchase cost – contract sale saving
+ transmission wheeling
+ energy / AS / fuel / capacity market purchase cost
– energy / AS / fuel / capacity market sale revenue
Subject to
• Energy balance constr aints
• Operation reserve constraints
• Generator and contr act chronological constraints: ramp, min up /down, min capacit y, et c.

• Generator and contract energy limits: hourly / daily / weekly / …
• Transmission limits
• Fuel limits: pipeline, daily / weekly/ …
• Emission limits: daily / weekly / …
• Others
The PLEXOS MIP algorithm solves the Portfolio Optimization problem by simultaneously obtaining
optimal solutions to unit commitment, economic dispatch, contract exercise, and market activates. This
approach produces what is called “ co-optimization” solutions.
Combined Cycle Generator Modeling
Due to the nature of the MIP formulation, the combined cycle generator can be modelled precisely using
its components, i.e., gas turbine generators and steam turbine generator. For example, the following
chart illustrates the 2x1 combined cy cle generator representation in PLEX OS.
Each component, the gas turbine generators and the steam turbine generator, is modelled as a regular
generator. The red line in the chart represents the waste heat transfer from the gas turbines to the Heat
Steam Recovery Generator (HSRG), and is modelled as a constraint in the PLEXOS MIP formulation.
Then, the PLEXOS unit commitment and economic dispatch algorithm can be used to determine the best
operational mode t o minimize the portf olio production cost of this sy stem.
PLEXOS® LT-PLAN - Long-term Capacity Expansion
Long-term (LT ) Capacity Expansion determines optimal investment decisions over long period of time,
usually up to 30 years. The PLEXOS LT-PLAN module accomplishes this by minimizing the Net Present
Value of forward-looking investment costs and the portfolio production cost. Therefore, the portfolio
cost minimization problem is expanded to include the investment cost and the investment-related
constraints as f ollows:
Minimize Portfolio Production Cost + In vestment Cost S ubject to
• Portfolio operation constraints
• Investment Constraints

Here, Investment Cost may include cost of new generator builds, cost of transmission expansion, and/or
cost of generator retirements. The Investment Constraints may include regional capacity reserve
margins, resource addition and retirement candidates (i.e. maximum units built / # retired), technical
and financial life spans, technology / fuel mix rules, Renewable Portfolio standard (RPS), etc. The build
and retirement candidates might include thermal generators, geothermal generators, hydro or pumped
storage hydro generators, wind or solar generators, transaction and demand side participation,
transmission augmentations, or generator retrofits.
The PLEXOS LT-PLAN solution is illustrated by the following chart. The x-axis corresponds to the
investment and the y-axis corresponds to the cost. As the investment increases, the production cost,
P(x), decreases (blue line) and the capital cost, C(x) increases (green line). The total cost (red line) is the
sum of the capital and production costs, C(x) + P(x). The PLEXOS LT–PLAN simulation returns the optimal
investment decision (x*) while observing the investment and operational constraints.
Total Cost C(x) + P(x)
Optimal Investment x*
Capital Cost C(x)
Investment x
Cost ($)
Production Cost P(x)
Stochastic Simulations to Incorporate Risk
In the portfolio optimization, risk is an important consideration in the corporate decision-making
environment. To incorporate risk, PLEXOS uses stochastic simulation to produce solutions with a statistic
distribution for a given set of the stochastic drivers, such as fuel prices, load forecast, renewable
generation profiles, market prices, etc. These stochastic drivers can be specified in PLEXOS
exogenously, i.e., the user provides the stochastic time series, or endogenously, i.e., PLEXOS produces
the stochastic time series for the stochastic drivers based on parameters provided by the user. For
example, the following chart shows a weekly gas price distribution over one hundred samples.

Here, the first four bars represent the maximum value, minimum value, standard deviation, and mean
value over the one hundred gas price samples. The remaining bars represent the one hundred
samples.
For the given stochastic driver time series, PLEXOS performs multiple iterations of simulations and
produces multiple iterations of solutions. The following histogram shows the distribution of portfolio
production costs for a week over the stochastic drivers of gas price, load forecast and market price
forecast.
The chart shows that the mean production cost in this week is about $180 million with a worst case
scenario of $260 million and a best case scenario of $161 million. The distribution of the weekly
production cost provides value insight to the corporate decision makers.
Testimonial
“Wärtsilä has used PLEXOS successfully for several years now. Purpose of our modelling work is to
quantify the value of flexibility for power systems and for individual customer portfolios. The studies
have clearly identified the need of more flexible generating resources in the future. Also, Energy
Exemplar’s willingness to develop the software, based on our needs, is highly appreciated”.
Ville Rimali
Wärtsilä
Europe

Reliability Analysis
The PLEXOS simulator is a powerful tool for performing reliability studies on electric power systems. The
simulator can calculate the standard metrics of LOLP, LOLE, EDNS and EENS from the PASA simulation
phase using convolution. However, PLEXOS can also use the detailed chronological simulation of ST
Schedule to produce the same metrics via Monte Carlo.
Energy Exemplar adopted this approach in the valuation of Loss of Load Expectation (LOLE) for the
market operated by ISO-New England, which covers the six states of New England in the north-eastern
US. The following describes this approach we used with PLEXOS.
How to setup PLEXOS® for Reliability via Monte Carlo
In a reliability study we are only interested in whether or not the system can meet the electric load, and
possibly ancillary services requirements, given all generation technical, transmission, fuel availability
and other user-defined constraints. The dispatch order of generators and the system marginal price is
not useful and thus we can simplify the simulation by removing all costs apart from the pseudo cost of
unserved energy (Value of Lost Load or VoLL). To do so, simply mark all cost information with a Scenario
and exclude that Scenario from your reliability Model run. Example of cost parameters include:
• Generator [Heat R ate] and [Load P oint]
• Generator [ VO&M Charge]
• Fuel [Price]
• Generator [Offer Price]
Having done this, the simulation will be less complex, with the objective function containing only costs
of unserved energy, dump energy and perhaps reserve shortage.
Loss of Load Expectation (LOLE)
Normally we compute the Loss of load probability (LOLP) as a reliability metric. This is a measure of the
probability that demand will exceed the capacity of the system in a given period and is expressed as a
percentage. As noted above, the periods when demand exceeds supply is a shortage event. An
alternative metric is the Loss of Load Expectation (LOLE), which is the number of days of outage or the
number of times in a given period that the load will be greater than the demand. This can be expressed
as LOLE = LOLP * N/24. A value of 1.0 LOLE over a single year signifies multiple periods or hours over
the year in which there were shortages. It is unlikely these shortages occurred in a single day but likely
over multiple hours over several or more days.
We used the Monte Carlo feature in PLEXOS to calculate the probability of the LOLE for each load
probability forecast in the New England. The first calculation was to benchmark the ISO-New England
Installed Capacity Requirement (“ICR”) calculation for the study period, or in this case the 2017 Peak
Load Forecast. The following demand forecast from ISO-NE’s ICR report (see Table 1 below) was used
for this analysis.

Table 1
We then performed a LOLE calculation with an average load-weighted risk approach. We prepared a
range of 16 different installed capacity levels in the ISO-NE control area and for each capacity level
evaluated the average load weighted LOLE at each of the above 10 load probabilities in Table 1.
The following figure, shows the demand distribution on the y-axis and the capacity distribution on the
x-axis. There are a total of 160 capacity and demand pairs in the following chart for which we
computed the average load weighted LOLE for using PLEXOS.
Figure 1
Then we prepared an average load weighted LOLE calculation for the 10 load probabilities for each
capacity point and derived 16 average load weighted LOLE’s as a function of installed capacity for
the ISO-NE control area. We derived a value of LOLE at the ISO-NE Net Installed Capacity
Requirement (NICR) value of 0.1 day per year which is consistent with reliability criteria for NICR.
With the 16 LOLE calculations we were also able to compute the cost of lost load at each installed
capacity level using the PLEXOS 8,760 hour time series simulation. PLEXOS estimates Unserved
Energy (USE) at each hourly interval if demand is greater than generation, or a shortage event.

This can happen primarily in summer months with reduced thermal generator capacities in
combination with load risk and forced generator outages. We then approximated the Value of Lost
Load (VoLL). For this analysis, we assumed a VoLL $20,000/MWh. Figure 2, below, displays an
example time series path of one of the load weighted risk replications for the 8760 hour simulation
performed. We charted the cost of lost load in millions of dollars, which is the result of multiplication
of the USE at each interval as described above by the value of VoLL selected above.
Figure 2
In Figure 2, the show an example of the 8760-hour simulation path for the peak demand and installed
capacity pair with the cost of unserved energy is displayed on an hour by hour basis in millions of
dollars. We ran 160 of these paths using Monte Carlo simulation and then took an average load
weighted cost of lost load across the demand risk for each of the 16 installed capacity values.

Figure 3, below displays a curve of the LOLE calculations as a function of percent of Net Installed
capacity Requirement (“NICR”) for the New England Control Area as well as the cost of lost load as a
function of percent NICR.
Figure 3
Figure 3 shows that the system LOLE reliability metric degrades rapidly for installed capacity values
below NICR. Also, the cost of lost load increases rapidly below NICR.
Testimonial
"I have been using PLEXOS to study the impact of wind and solar variability on power system reliability,
both for thermal and hydrothermal systems. PLEXOS makes quite easy to run Monte Carlo simulations
and to define different modelling scenarios. Its flexibility has allowed us to evaluate how operational
aspects such as solar and wind profiles, transmission constraints, hydro inflows, operational reserve
levels, and energy storage may impact the adequacy contribution of variable generation."
Professor Esteban Gil
Universidad Federico Santa Maria
South America

Capacity Expansion Planning
Flexible power generation and interconnection capacity
needs of the Italian power system using PLEXOS® LT plan
One of the main developments of the last few years on the Italian power market has been the significant
increase of renewable capacity in the form of wind and solar power plants. Using a PLEXOS dataset
based on data available from Terna and various utilities, we have modelled the Italian power market
through the six geographical zones and considered the medium term expected scenarios. The Southern
zones (namely South, Sicily and Sardinia), already characterised by a significant amount of renewable
capacity, are expected to experience the largest concentration of renewable development. In addition,
these regions are characterised by a relative lower level of interconnection with neighboring countries
compared to the Northern zones. Given the variability introduced on the system by the increasing
renewable capacity and the relative isolation of the Southern zones, the deployment of flexible power
plants and the expansion of interconnection capacity are assessed using the PLEXOS Long Term (LT) Plan
feature. This feature, designed for long term optimal investment analysis, provides useful insights about
potential bottlenecks, system flexibility and power prices, and is a critical tool for analysts, researchers,
regulatory and utility decision makers to evaluate network adequacy and expansion, security of supply,
investment options and energy policy development.
Introduction and Model Description
The power market simulation and analysis software PLEXOS® Integrated Energy Model has been used to
represent the Italian power market in 2012 and its development in the medium term, considered up to
the year 2017. Energy Exemplar built a dataset including all the relevant plants within each
geographical market zone in 2012, with the poles of limited production aggregated to the surrounding
geographical zones as: BRNN, FOGN and ROS to the SUD zone, and PRGP to the SICI zone.
Figure 1: Geographical zones of the Italian power market.
Installed capacities of conventional plants and aggregated figures of renewable generators are
obtained from Terna reports, utilities websites and ENTSO’s Scenario Outlook & Adequacy Forecasts
reports. Such capacities cover the year 2012 and extend up to 2018. In addition, for each technology
we have taken a representative set of operational characteristics (heat rates, ramp rates, start costs etc.)
and applied it to each of the plants of that particular technology. Renewable generation profiles are
obtained from the in-built PLEXOS® feature to create stochastic samples from a given profile, taken as
the real 2012 data in this instance.

The long term country demand profiles, and zonal profiles for Italy, were built using the 2012 hourly
values and the PLEXOS in-built load forecasting feature that uses a given base profile, annual peak and
energy demand forecasts; such a procedure was applied to each of the countries included in the model
(Italy, France, Switzerland, Austria, Slovenia, Albania, Montenegro and Greece).
The aim of this study was to assess the need for additional interconnector capacity or the development
of flexible power plants, and for that reason we employed PLEXOS® Long Term (LT) Plan feature.
Long Term Capacity Expansion Planning with PLEXOS®
Assessing the risk and return of new investments both in new power plants and in the reinforcement of
the transmission grid requires highly specialized tools. PLEXOS® Integrated Energy Model feature of
Capacity Expansion refers to the problem of finding the optimal combination of generation new builds
(and retirements) and transmission upgrades (and retirements) that minimizes the net present value
(NPV) of the total costs of the system over a long-term planning horizon. That is, to simultaneously solve
a generation and transmission capacity expansion problem and a dispatch problem from a central
planning, long-term perspective.
The optimization problem takes into account in the objective function two types of costs; capital costs
(costs of new generator builds and transmission expansion costs) and production costs (costs of
operating the system with any given set of existing and new builds and transmission network as well as
the notional cost of unserved energy). The PLEXOS LT Plan phase by default spans the whole simulation
planning horizon in a single optimization 'step', while it gives the user the option to control the type of
chronology used, and choose between preserving the full chronology (Fitted chronology) or using the
partial chronology (Load duration curves, LDC).
Results and Discussion
In this case study we define generator and transmission expansion candidates which PLEXOS will assess
in order to give the optimal solution. Therefore, in the inputs we define the optional investments to be
either the expansion of the existing interconnectors (with France, Switzerland, Slovenia and Greece) or
the building of new ones (with Croatia, Montenegro and Albania), as per Terna development plans. In
addition, we give PLEXOS the option to select to build new CCGT, GT or ST power plants in all of the 6
Italian zones.
The optimal solution as calculated by the PLEXOS LT Plan shows that between 2015 and 2017 all the
existing interconnectors used to import electricity into the country are congested most of the time (see
Figure 2); the list includes those with France, Switzerland, Austria and Slovenia and excludes only the
one with Greece, which is operating mainly for export.
Figure 2: Congestion of interconnectors with Northern neighbour countries

Given the need for more interconnector capacity, the PLEXOS solution is to build an interconnector
linking the South zone with Albania. This interconnector is indeed in the Terna plans and the PLEXOS
result is well suited, given that most of the renewable capacity and therefore the greatest need for
balance of intermittent generation is located in the South. Together with the security of supply
constraint, which is met by this new build, the other critical parameter largely impacted by the
availability of interconnection capacity is the price in the country. Given the limited horizon in the
PLEXOS model compared to the typical economic lifetime of an interconnector project, further builds are
prevented, as the cash-flow from the lower-cost imported electricity savings does not justify the
investment. The net result is a higher power price compared to the neighbour countries.
In addition to the interconnector with Albania, the PLEXOS optimization includes also the building of
new CCGT plants in Sicily. This result is justified by the limited spare capacity of conventional plants on
the island and represents a measure to prevent the prices to be too sensitive to the generation from
renewables, as shown in Figure 3. Nevertheless, the price in Sicily is higher than in the rest of Italy, as
shown in Figure 4, with significant peaks in the summer. The congestion of the interconnector with the
mainland for most of the year suggests that an increased capacity might limit the price spread, but the
cash-flow argument used for the interconnectors with neighbour countries applies also to the one
between Sicily and the mainland.
Figure 3: Effect of new-built CCGT plants on Sicilian power prices
Figure 4: Comparison of power prices among the 6 Italian zones
Another insight also obtained from the PLEXOS LT Plan results is the level of congestion of all of the
interconnectors exporting electricity from the South zone, as shown in Figure 5.

Figure 5: Congestion of interconnectors connected to the South zone
Conclusions
The Terna plans to increase these export capacities and the new interconnector with Albania described
above, is a way to address the large imbalance between demand and supply, as shown in Figure 6.
Such expansion would also avoid the significant cycling costs and low capacity factors of the
conventional power plants in the region, thus mitigating their effect on the energy prices in the area with
likely positive effects for all the country.
Figure 6: Demand and supply in the South zone
Testimonial
"Since I started working with PLEXOS, I have been surprised of how powerful and useful it is. PLEXOS
covers such extensive and accurate approach of modelling the market and provides different
possibilities of modelling the different factors and drivers of the power market. I used PLEXOS to model
different power markets across the world, every of which had their own particularities, which implies the
difficulty of different approach when modelling. Thanks to such a powerful tool, the required knowledge
and the incredibly great support from energy exemplar, every client case I used PLEXOS for the
modelling part, the client happened to be very satisfied with the results and my company as well. I
strongly recommend it to those new potential users thinking about joining the PLEXOS practice!”
Elena De Juan Salgado
Boston Consulting Group
Europe

Hydro Optimization
Power systems having both hydro-electric and thermal generation require a systematic and coordinated
approach in order to determine an optimal policy for dam operations. The goal of a hydro-thermal
planning tool is to minimize the expected thermal costs along the simulation period. These types of
problems generally require stochastic analysis to deal with inflow uncertainty. This can increase the
mathematical size of the problem and can easily become cumbersome to solve.
PLEXOS Integrated Energy Model offers many features to deal with the hydro-thermal coordination
problem. In addition, it offers a seamless integration of phases, making it possible to determine an
optimal planning solution in the medium-term and then use the obtained results in a detailed short-term
unit commitment and economic dispatch (UCED) problem with increased granularity. For example,
weekly targets as constraints filter down to produce hourly electricity spot prices.
The Challenge
The systematic coordination of a system composed of both hydro-electric and thermal plants requires
determining an operational strategy that for each stage of the planning horizon produces a scheduling
plan of generation. This strategy minimizes the expected operational cost along the period, which is
mainly composed of fuel costs plus penalties for failure in load supply. The problem becomes complex
to solve because generally in hydro systems:
• Natural inflows ( by nature) are stochastic processes.
• Availability of w ater stored in dams is limit ed.
• There are complex cascading h ydro systems.
• Water usage policies and en vironmental releases such as irrigation settlement s.
Water as a fuel supply is cost-free, but its opportunity cost is fundamental to finding the optimal strategy
for operations. This issue creates the need for a decision in a given time period. Storages can’t be
drained too low, which might incur generation shortfalls or excessive thermal output. On the other hand,
we also want to avoid spillage of water and lost generation opportunities. Figure 1 summarizes the
dilemma a hydro power planner faces to operate a dam.
Figure 1: Diagram showing the dilemma hydro power planner faces under uncertainty

The Solution
PLEXOS can find a hydro releasing policy that minimizes the expected thermal cost by formulating an
optimization problem such as the following:
Min {Variable Costs}
Subject to:
Energy Balance Equation
FlowLimits
GenerationLimits
Hydro Balance: 1
The hydro balance equation shows the link between decisions in both the present and future.
Since water is free (no fuel cost) it is necessary to specify a final condition so as to minimise thermal costs
along the simulation period, and to avoid the storage being completely drained. These final conditions
can be represented as a target or a proxy for opportunity costs such as a deviation from targets, usually
known as the future cost function or scrap value function.
To handle uncertainty in the inflows, PLEXOS offers stochastic optimization techniques in both 2-stage
and multistage. Both techniques minimize the expected cost (or maximum benefits) of the system.
PLEXOS® Features Explored
MT and ST schedule
To determine optimal targets or deviation targets, PLEXOS can analyse a longer horizon using Medium
Term (MT) analysis and if the mathematical size of the problem is unsolvable a reduction in granularity
(called demand blocks) can be applied to produce a mathematical problem that is faster to solve. Then
the targets or deviation targets can be calculated from this optimal solution. This simplification produces
an end volume at each block, then the MT problem passes the target or future cost function to each short
term (ST) step so as to obtain a higher resolution release policy. Figure 2 shows an example of MT target
(red line) and ST end volumes (blue area).

Cascading Hydro with Storage and Waterways
To create models of cascading hydro networks a combination of Storage, Waterway, and Generator
objects can be used. PLEXOS offers the capability to model a cascading hydroelectric system like the
one shown in Figure 3.
Figure 3: Schematic representation of Storages and Waterways
Storage objects are used to represent storage reservoirs with any storage capability or even simple
junctions in a river-chain. Each storage can connect to one or more generators or waterways to create
a model of a river chain.
Waterway objects connect the storages or spill water from the storage 'to the sea'. Waterway flows can
delay the water flow, can have bounds placed in their flows and can limit the rate of change in their
flows. The Constraint class makes it possible to relate water flow rates and generator efficiencies to
storage elevations.
Pumped Storages Power Plants
Pumped storage plants store energy in the form of water. Pumps move water from a lower elevation
reservoir (the 'tail' storage) to a higher elevation ('head' storage). To optimise the operation of a
pumped storage power plant, PLEXOS formulates an optimization problem to decide when to release
and pump water in order to minimize costs (or maximize benefits). In general one should expect that
low-cost off-peak electric power is used to run the pumps and during periods of high electric demand
and high price, the stored water is released through turbines to generate power.
Figure 4: Pumped Storage cycle. Generation left hand side and pumping right hand side

Stochastic Optimization
PLEXOS can find the best outcome given the uncertainty in inflows. The problem is summarized in Figure
1 and can be solved using a stochastic optimization formulation. PLEXOS offers a two stages and
multistage optimization approach. The difference is that multistage techniques allows a re-evaluation of
the policy after some time, the period when this happens is called a “Stage”. In a two-stage approach,
the optimal policy is never re-evaluated because the information is not revealed at an intermediate
stage during the simulation.
These two types of problems are represented in the below scenario trees.
Root 1st Period 2nd Period 2nd Stage Root
1st Stage
1st Period
2nd Stage
2nd Period
Figure 5: Left hand side represent 2 Stages SO, right hand side represents multi stage SO
Multi Stage Tree Reduction
When using multistage stochastic optimization, a large number of possible scenarios can be generated.
For such a high number of scenarios, it is impossible to numerically obtain a solution for the multi-stage
optimisation problem.
...
Root Stage 1 Stage 2 Stage 3 Stage T
That number of scenarios is equal to:
Branches_perStage(NStages)
Example: 21 stages and 3 branches per node
321 = 1.04604 x 1010 Scenarios!
This problem is impossible to solve
Figure 6: Multi Stage dimensionality issue
To help solve this problem, PLEXOS offers scenario reduction techniques. These techniques use
strategies to reduce the number of scenarios in the optimization problem using algorithms for
constructing a multi-stage scenario tree out of a given set of scenarios.
Since generating a very small number of scenarios by Monte Carlo Simulations is not desired because
less scenarios give less information, the objective is to lose minimum information by the reduction
process applied to the whole set of scenarios.

Case Study Results and Discussion
The case study summarized in Figure 7 illustrates the application of stochastic techniques in the hydro
planning problem. The power system under study consists of 3 Power plants and there are 12 possible
inflow sequences. The objective is to find an optimal trajectory for the storage along 1 year.
Max Cap = 40,000 MW
Hydro Hydro
Initial Volume = 10600 GWh
Max Volume = 106200 GWh
Min Volume = 1000 GWh
Max Cap = 20,000 MW
Coal
Nuclear
Max Cap = 12,000 MW
Load
Figure 7: Left hand side schematic of the Power System under study, Right hand side load profile
Three simulations are performed: Deterministic, 2 Stage and Multistage. The optimal trajectories
obtained are summarized in Figure 8.
Figure 8: Comparison of Trajectories for deterministic, 2 stages and multistage (4 stages).
2 stages obtains one single trajectory as there is no chance to re-evaluate the decision through time. On
the other hand, multistage perform a re-evaluation at each stage, therefore producing a set of optimal
past dependant trajectories. The deterministic approach gives optimal trajectories for each sample but
it is not capable of answering the question “what decision do I have to make now given the uncertainty
in the inputs?”
To evaluate the quality of the policies obtained, we can randomly select one of the policies, run the 12
inflow possibilities and calculate the expected costs. The results are summarised in Table 1.
Deterministic (Fixed Targets)
Expected Costs (Million $) 42,612
Table 1: Summary of results
2 Stages
21,638
Multistage
20,329
The results show that multistage stochastic optimization manages the storage better. Also 2 stages gives
lower expected generation costs compared to the case with deterministic fixed trajectories.
Conclusions
Hydro – thermal power system modelling has lots of challenges and optimization techniques are
necessary to provide the required answers. PLEXOS features can solve hydro – thermal coordination
problems to ensure the user minimizes the cost or alternatively maximizes the benefits of a
hydro-thermal portfolio. PLEXOS offers the flexibility to customise the stochastic resultant problem,
providing the user with a powerful tool to manage hydro uncertainty that can be integrated with the rest
of PLEXOS features.

Transmission Analysis
SAPP establishment and development
The Southern African Power Pool (SAPP) was formed in 1995 to promote trading of electricity between
nations in the Southern African Development Community (SADC). Initially, national utilities of each
SADC country were included to co-operatively trade electricity (via bilateral contracts and a Short Term
Energy market (STEM)). The STEM was replaced in 2009 with a competitive market (the Day Ahead
Market (DAM)) and inclusion of Independent Power Producers (IPPs) and Independent Transmission
Companies (ITCs) was later allowed.
The SAPP membership of 16 (Table 1, as of 2013) is shown in a-Tanzania-Kenya Interconnector., the
majority of SAPP members are interconnected with the only isolated members being ESCOM (Malawi),
ENE (Angola) and TANESCO (Tanzania).
Entity
Botswana Power Corporation
Electricidade de Mocambique
Hidro Electrica Cahora Bassa
Mozambique Transmission Company
Electricity Supply Corporation of Malawi
Empresa Nacional de Electridade
Eskom
Lesotho Electricity Corporation
NamPower
Societe Nationale d’Electricite
Swaziland Electricity Company
Tanzania Electricity Supply Company
ZESCO Ltd
Copperbelt Energy Corporation
Lunsemfwa Hydro Power Company
Zimbabwe Electricity Supply Authority
OP = Operating Member
NP = Non-Operating Member
OB = Observer
ITC = Independent Transmission Company
Status
OP
OP
OB
OB
NP
NP
OP
OP
OP
OP
OP
NP
OP
ITC
IPP
OP
Abbreviation
BPC
EdM
HCB
MOTRACO
ESCOM
ENE
Eskom
LEC
NamPower
SNEL
SEC
TANESCO
ZESCO
CEC
LHPC
ZESA
Country
Botswana
Mozambique
Mozambique
Mozambique
Malawi
Angola
South Africa
Lesotho
Namibia
DRC
Swaziland
Tanzania
Zambia
Zambia
Zambia
Zimbabwe
Table 1: SAPP membership (2013)
A constrained SAPP
The interconnected SAPP has been significantly constrained for generation capacity in the last decade.
SAPP demand (incl. typical 15% reserve margin) for compared to net SAPP capacity (with/without South
Africa) has been consistently higher since 2006. This trend is magnified when excluding the dominant
player in the SAPP (Eskom, South Africa). Although, there are a number of rehabilitation and new
generation projects currently underway to remove these constraints in the medium term (mostly in
Botswana, Mozambique, South Africa, Zambia and Zimbabwe).

In addition to existing generation capacity constraints, transmission capacity constraints have had a
significant effect on the volume of trading within SAPP (whether via bilateral contracts or DAM). For
2013/2014, on average, 70% of energy matched in the DAM was not traded and 77% of bilateral
contracts granted were not traded.
Again, there are projects in the pipeline that to rectify these transmission constraints including the
ZIZABONA project and the Central Transmission Corridor project. Other notable priority SAPP
transmission projects include the interconnection of NP members i.e. Namibia-Angola Interconnector,
Mozambique-Malawi Interconnector and Zambia-Tanzania-Kenya Interconnector.
SAPP modelling with PLEXOS®
A sufficiently detailed regional transmission model of the SAPP in the short term, medium term and long
term will allow existing national utilities and regulators as well as prospective IPPs and ITCs in the SAPP
to fully assess the risks and opportunities present. This is where a software tool like the PLEXOS®
Integrated Energy Model (“PLEXOS®”) can assist.
Based on only public domain information and generally accepted industry practice, a regional level
SAPP transmission model has been developed in PLEXOS. This dataset includes a representation of each
SAPP country (including load profiles and hourly demand and energy forecasts for 2010-2025), existing
transmission interconnectors, existing major generators and country specific fuels.
The transmission components of the PLEXOS SAPP dataset are shown geographically in Figure 1
(extracted directly from PLEXOS mapping interface) with the SAPP interconnectors summarised in Table
2.
From
Botswana
Botswana
Botswana
Botswana
DRC
DRC
Lesotho
Mozambique (Central)
Mozambique (South)
Mozambique (South)
Mozambique (South)
Mozambique (South)
Mozambique (North)
Mozambique (Central)
Namibia
Namibia
South Africa
South Africa
Zambia
Zambia
To
South Africa
South Africa
Zimbabwe
Zimbabwe
Zambia
Zambia
South Africa
Mozambique (North)
South Africa
South Africa
South Africa
Swaziland
Zimbabwe
Zimbabwe
South Africa
South Africa
Swaziland
Zombabwe
Namibia
Zimbabwe
Voltage Level (kV)
400
132
400
220
330
500
132
110
533
110
400
400
330
110
400
220
400
132
350
330
Transfer Capacity (MW)
190
425
220
205
247
560
90
30
1920
67
1100
1000
220
38
380
195
1100
15
180
428
Table 2: Existing SAPP interconnectors modelled in PLEXOS dataset

Figure 1: Existing SAPP transmission model in PLEXOS (geographical)
Interconnectors (the central corridor example)
The developed SAPP dataset was used to assess three case studies:
• 2012 SAPP as -is (Existing S APP system with e xisting bilateral contracts)
• 2012: SAPP Free (existing SAPP system ignoring bilat eral contracts)
• 2012: SAPP Unconstrained (existing SAPP with no tr ansmission constraints)
The central corridor (Botswana-Zimbabwe-Zambia) is known to be a significant transmission constraint.
This can be seen in Figure 2 where a typical days’ power-flows taken from the PLEXOS SAPP model are
shown. As can be seen, when free trade is allowed, the South Africa-Botswana interconnector gets
pushed to its maximum capacity. Some of this power is used in Botswana but a considerable amount is
wheeled via the Botswana-Zimbabwe interconnector further north especially when no transmission
constraints are imposed.
Figure 2: SAPP central corridor power-flows over a typical day (SA-Botswana)

Conclusions
PLEXOS has been used to develop a regional level transmission model of the SAPP. This model includes
all major generators, loads and load profiles, transmission interconnectors, storages for pumped
storage generators, country specific generator fuels and CO2 emissions from generators.
The developed PLEXOS model was used to assess the impact of existing transmission constraints and
bilateral contracts in SAPP. These case studies revealed that existing energy imbalances in SAPP
countries can be significantly reduced if interconnectors are freed up to allow for the most optimal trade
of electricity between SAPP nations. In addition, in a scenario where all SAPP nations are assumed to
be interconnected and no transmission constraints exist, energy imbalances are removed completely.
Testimonial
“It is truly amazing how Energy Exemplar seem to be breaking new boundaries. Firstly, the introduction
of chronological optimization for the Long Term Optimization studies seems to be paced right at time
where renewables are the “flavour of the moment”. Secondly the introduction of the gas module again
pitched at the time where shale gas is shaping the market. Not only are the new features paced
appropriately, the algorithms and results are truly stunning and this will definitely help to put policy
options on a scientific footing. PLEXOS has been fascinating to use to produce the South African
Integrated Resource Plan (IRP), the stochastic optimization aspect of it was helpful to establish the
position of least regret. We are more confident that the next IRP will be more informative given that
PLEXOS has given us new questions which need answers...”
Sipho Mdhluli
ESKOM
Africa

Gas System Planning
These are interesting times for the Eastern Australian gas market with the impending LNG play coming
online. The previously stable and long-term contract market for domestic gas supply on the east coast
will be subject to market forces that are in part determined on the global stage. Exactly how the market
will respond to these changes is not yet clear, however, one important question on the mind of every
market participant is: “do we have enough gas to meet medium to long term domestic demand?” In
answering this question and given the opaque nature of the gas market, there exists an asymmetry of
opinions in the demand – supply market analysis.
Historically, the Eastern Australian gas market has been relatively insulated from international market
pressures, operating on a somewhat “self-sufficient” business model. However, the recent large scale
development of export LNG facilities is about to change all of that as the eastern Australian gas market
has found itself in the midst of the largest structural revolution since the privatisation of gas infrastructure
assets in the 1990s. Predictably, participants in the eastern gas market are keen to understand the
effects this impending LNG play will have not only on their market positions, but also on the whole
production – demand supply chain on the domestic front. In order to predict any effect the LNG
development will have on the domestic market, a comprehensive understanding of the domestic supply
chain is necessary.
This paper aims to analyse the demand - supply interactions of the eastern Australian gas market, in
order to establish any potential risk of gas shortage in the medium to long term. Employing a model
developed through the new gas module from PLEXOS, an analysis of the market is carried out from 2013
- 2024. An examination of this nature is fundamental to understanding not only the LNG effects on the
market, but also to identify any potential supply shortfalls looming in the near future.
Gas Model Description
Scheduling of gas energy markets require highly specialized tools to conquer the complexity of this
dynamic, commercial and regulatory landscape. PLEXOS integrated gas and electricity simulation
software package provides technical and mathematical solutions to meet planning needs. The gas
module in PLEXOS allows detailed modelling of the physical delivery of gas from producing fields,
through pipelines and storages (including linepack), to demand points with the capability to model any
physical constraint along the supply chain. Furthermore, the PLEXOS® Integrated Energy Model has the
ability to optimize gas and electricity markets by simultaneously solving both markets’ parameters,
allowing decision makers to trade-off gas investments, constraints and costs against other alternatives.
The following elements are the main components of the gas network modelled in the present study:

Model Inputs
Icon Class Description
Gas Field Field from which gas is extracted.
Gas Storage Storage where gas can be injected and extracted.
Gas Pipeline Pipeline for transporting gas.
Gas Node Connection point in gas network.
Gas Demand Demand for gas covering one or more nodes.
The model developed for this report is built with information available in the National Gas Market
Bulletin and the Australia Energy Market Operator (AEMO). The demand profiles for 2013 used for the
demand zones illustrated in Figure 1 are obtained from the National Gas Market Bulletin.
Figure 1: Gas network modelled in PLEXOS, the main demand zones included in this study are:
Mount Isa, Gladstone, Brisbane, Adelaide, Sydney, Melbourne and Hobart
Using the projections supplied by AEMO in the Gas Statement of Opportunities (GSOO) 2013, a
demand forecast is built as depicted in Figure 2 showing projections until 2024 for the major demand
centres.
To accommodate the impact on gas production from the LNG export demand, we run a sensitivity study
on the eastern Australian gas market with the production from the Bowen Surat gas fields (largest known
and proven CSG basin) turned off and completely unavailable for domestic market. This assumption
implies that all the production from this reserve is for export only therefore left out of this study (domestic
demand-supply model). In addition, pipeline outage and maintenances are not included, and all
pipelines are assumed to be online for the next 12 years. We also assume that the NSW CSG
developments go ahead and are all available from the start of the planning horizon.

Results and discussions
Firstly, we run a “base” model of the eastern Australian gas market for 12 years with all the producing
fields available for production. As expected, we see no gas shortages in the system with some CSG gas
fields not even required to come online in this time.
Figure 2: Demand outlook from 2013 - 2024 as projected by the GSOO 2013
As a sensitivity to model the LNG effects on the demand – supply dynamics on the eastern domestic
market, we turn off and completely make production from the Bowen Surat gas fields (the largest proven
producing CSG reserves in eastern Australia) unavailable for domestic demand. A justification for this
sensitivity is the assumption that all the gas produced from this basin for the next 12 years are fully
contracted to LNG export. From the results, we observe that there is enough gas in the eastern reserves
to accommodate the impending LNG play as there are no shortages in any of the demand zones
considered in this study. In addition, we observe that CSG fields which were offline in the “base” model
came online producing at almost full capacity to offset the ‘no production’ scenario from the Bowen
Surat gas fields.

Figure 3: Gas production with Bowen-Surat reserve completely contracted to LNG export from
2013 - 2024
This result is consistent with other studies in the literature, reaffirming the notion that eastern Australia
has enough gas in its reserves to accommodate LNG exports at the current production rates and demand
outlook for the next decade.
Final comments
Given the level of uncertainty surrounding the east coast gas market, there is a need for a sophisticated
and complex analysis of the different scenarios and sensitivities surrounding the current market. As such,
some open questions arising from this study, all of which can be addressed with some further modelling
employing PLEXOS, include:
• Will increasing LNG e xports create a domestic short age and t o what e xtent will this be seen in the
short term?
• How does the gas landscape change with the intr oduction of a new carbon scheme in futur e years?
• How will the domestic gas mark et react to a netback of gas price t o international LNG prices?
• Will a shift in the domestic demand outlook over the ne xt decade impact the suppl y - demand
dynamics of the market?
Testimonial
“PLEXOS is our tool of choice for NEM market & gas modelling. It gives us the flexibility to set up the
model as we choose, and its scenario management is very sophisticated, allowing consistent
assumptions to be applied to short, medium or long-term outlooks”.
Ben Howard
AGL
Asia-Pacific

Renewable Generation Integration
The uncertainty and variability nature of the renewable generation penetration creates integration
challenges within the electric industry from the portfolio scale to the regional ISO level. The challenges
include:
1. Is the Power Market or the portfolio ready for the renewable generation variability and uncertainty?
2. What can be done to be prepared for the renewable generation?
3. What is the cost for this readiness?
4. Any changes are needed for the power market or portfolio operational procedure?
5. Can new market products improve the readiness?
6. Intra-hourly inter-BAA generation and capacity sharing?
7. All questions are leading toward a better solution – Energy Imbalance Market (EIM).
This article first reviews the uncertainty and variability of the renewable generation; then briefly
describes the approach to prepare the system for the renewable generation; third, the EIM is presented
for the ramp capacity sharing between the Balancing Authority Areas (BAA) to improve the system
readiness for the greater renewable generation penetration.
The Uncertainty and Variability of Renewable Generation
Here the renewable generation usually refers to the wind miller generation and solar generation. The
following two tables shows the 5-min actual solar and wind generation and Day-ahead (DA) and
Hour-ahead (HA) forecasts for a 130 MW solar project and a 220 MW wind project.

The following histograms shows the solar and wind generation forecast error distributions.
The DA and HA unit commitment will be performed based on the DA and HA solar and wind generation
forecasts. Especially, the DA unit commitment may be quite off from the actual solar and wind
generation.

What can be done to prepare for the Renewable Generation?
To cover the uncertainty and variability of the renewable generation, additional reserves can be
introduced during the DA and HA unit commitment. The additional reserves are usually called
“Flexibility Up” and “Flexibility Down” reserves. For the detail description of the reserves for the
renewable generation, please refer the article.
The flexibility up and down reserve requirements are determined by the statistics of the renewable
generation forecast errors. The flexibility up and down reserves are provided by the generator ramp
capacity. However to quantify the system readiness is best through the sub-hourly simulations with the
flexibility up and down deployment.
Energy Exemplar performs a few renewable generation integration studies using the 3-stage DA-HA-RT
sequential simulation approach. This approach can be illustrated in the following flow-chart.
Hourly DA SCUC/ED
Simulation in 24 hours
24-hour Unit
Commitment
Schedules for
Long-starts
Hourly DA SCUC/ED
Simulation with five hours
look-ahead
Unit Commitment
for all generators
5-min RT SCED Simulation
with 5-min look ahead
DA Forecasted
Load/Wind/Solar
HA Forecasted
Load/Wind/Solar
Sub-hourly “Actual”
Load/Wind/Solar
Contingency, Flex and
Regulation Reserve
Regulation Reserve
Regulation Reserve
The 3-stage DA-HA-RT sequential simulation approach is described as follows.
• PLEXOS DA simulation mimics the DA Security Constrained Unit Commitment and Economic
Dispatch (SCUC/SCED)
- Day-ahead forecasted load/wind/solar generation time series are used;
- Hourly simulation interval;
- The SCUC/ED optimization window is one day plus a few hour look-ahead;
- The transmission network is modelled at the zonal or nodal level;
- The contingency, DA flexibility / regulation up and down reserves are modelled.
• PLEXOS HA simulation mimics the intra-day or Hour-ahead SCUC/SCED
- 3-hour-ahead forecasted load/wind/solar generation time series are used;
- Hourly simulation interval;
- The SCUC/ED optimization window is a few hours plus a few hour look-ahead;
- The contingency, 3-HA flexibility / regulation up and down reserves are modelled;
- The unit commitment patterns from the DA simulation for the long start up generators are frozen;

• PLEXOS RT simulation mimics the 5-min real-time SCED
- The “actual” 5-min load/wind/solar generation time series are used;
- 5-min simulation interval;
- The SCUC/ED optimization window is a few 5 minutes plus a few 5-min look-ahead;
- The contingency and regulation up / down reserves are modelled. However, the flexibility up and
down reserves are not modelled. The implication is that the capacity held in the 3-HA simulation
for the flexibility reserves is deployed to cover the load and renewable generation variability and
uncertainty at the 5-min interval;
- The unit commitment patterns from the DA simulation for all generators, except the peaking
generators, are frozen.
The solutions from the 5-min RT simulation will include the over-generation, un-served energy,
contingency and regulation reserve shortfall. Usually the non-zero values of these indices indicate the
system is not completely ready for the renewable generation or the inadequacy of the system ramp
capacity. In addition, the 3-stage Da-HA-RT sequential simulation approach can be used to quantify
the effectiveness of the system operation procedure evolution, new products, etc.
Energy Imbalance Market
One of the effective operation procedure is the Energy Imbalance Market (EIM). In a EIM, each BAA in
the EIM performs the DA and HA Security Constrained Unit Commitment and Dispatch to minimize its
production cost. However, at the sub-hourly level, the members of the EIM will provide the generation
ramp capacity in a form of sub-hourly bidding quantity and price. The generation ramp capacity will
cover the sub-hourly load and renewable generation uncertainty and variability in the EIM in an
economic manner. Therefore, EIM allows the generation and flexibility and regulation reserve provision
sharing at a sub-hourly interval to cover the renewable generation variability and uncertainty with the
minimum cost. For the details of the EIM studies and VGS studies, please refer to the project reports.
To evaluate the benefit of the EIM, the sub-hourly RT simulation in the 3-stage DA-HA-RT sequential
simulation is performed twice: the Business As Usual (BAU) and EIM. In the sub-hourly RT simulation for
the BAU case, the hourly interchange between the BAA’s from the HA simulation are frozen to mimic the
current BAA operation practice. In the sub-hourly RT simulation for the EIM case, the constraints of the
hourly interchange between the BAA’s is relaxed to allow the sub-hourly interchange among the
members of the EIM. The produce cost difference of these two simulations is the benefit of the EIM.
Testimonial
“For the last four years I used Energy Exemplar’s PLEXOS modelling tool. This has, for the first time,
allowed for the effects of increasing amounts of renewable non-synchronous generation to be analysed
in terms of costs, curtailment and system services availability in the future electricity systems of Western
Europe.”
Edward Mc Garrigle
Bord Gais Energy
Europe

L e a d i n g t h e f i e l d i n
E n e r g y M a r k e t M o d e l l i n g
Roseville
Hartford
London
Johannesburg
Clients Adelaide
Of f ices
Energy Exemplar Pty Ltd
Software Development Office
17 Bagot St
North Adelaide
SA 5006, Australia
+61 8 8361 9312
info@energyexemplar.com
Energy Exemplar LLC
West Coast USA Office
Suite 120, 3013 Douglas Blvd
Roseville, CA 95661
United States of America
+1 916 722 1484
westcoastusa@energyexemplar.com
Energy Exemplar Ltd
Europe Office
Building 3, Chiswick Park
566 Chiswick High Road
Chiswick, London, W4 5YA
United Kingdom
+44 208 899 6500
europe@energyexemplar.com
Energy Exemplar LLC
East Coast USA Office
20 Church Street, Ste. 790
Hartford, CT 06103
United States of America
+1 860 461 0761
eastcoastusa@energyexemplar.com
Energy Exemplar Pty Ltd
Africa Office
West Tower Nelson Mandela Square
Cnr 5th & Maude Street
Sandton, 2196
Johannesburg, South Africa
+27 11 881 5889
africa@energyexemplar.com
e n e r g y e x e m p l a r . c o m

PLEXOS applications

Recommended

Recommended

More Related Content

What's hot

What's hot (20)

Similar to PLEXOS applications

Similar to PLEXOS applications (20)

Recently uploaded

Recently uploaded (20)

PLEXOS applications