ELEC9715 Exam Preperation Questions

December 7, 2015

Abstract
This document provides some sample questions made up by students, to help you study
for the final exam. We provide no guarantees about the correctness of the answers, or
whether they closely reflect what you’ll see in the exam. If you find any mistakes, email
vicepresident@elsoc.net.

Topic 1 Week 3: Operation envolope of electricity industry

technology, variable operating cost and actual oper-
ation

Question

State primary energy forms in the NEM and typical levels of on-site storage. Explain how this
affects operation between fuel types

Answer

Primary energy forms in the NEM include coal (brown and black), gas, wind, solar (PV & thermal),
biomass, hydro, liquid fuel. Typical levels of storage are:

• Coal can have stockpiles from days to weeks,

• Gas can have linepack from hours to days,
• hydro depends on the level upstream reservoirs
• Liquid is hours to days
• Wind and PV have zero primary energy storage
• Solar thermal can have heat storage tanks currently from hours up to a day.

Coal, gas, liquid and to some extent hydro are considered dispatchable generation. That is they
can operate whenever required. Wind and PV are not considered dispatchable as their primary
fuel source is non-storable. This means their operation is simply to run when the fuel is available
and run as hard as possible. Current levels of non-dispatchable generation in the NEM make them
unreliable for baseload generation as they cannot be kept running when required. Dispatchable
generation has primary fuel storage and so can be run to capacity when the market requires it. Coal,
gas and liquid are fully dispatchable. Hydro is conditionally dispatchable due the level of storage
being dependant on rainfall and snowmelt in reservoir catchments. Dispatchable generators can

1
choose when to generate and thus are able to employ a much greater range of operating strategies
compared to non-dispatchable.

Solar thermal technology can vary due to the heat stored in the working fluid and the option to
have heat storage tanks. Heat in the working fluid allows partial mitigation of temporary loss of
its primary energy source (the sun), making operation more predictable than PV and more stable
than wind. Heat storage tanks allows the plant to be dispatchable up to the level of the energy
capacity of the tanks.

Additionally:

• Solar thermal with storage can be used to store excess energy from “spilling” of RE or when
market prices are low to sell back at a higher price. This is because they have electric heaters
in the tanks to prevent fluids like molten salt from solidifying. 400MW biogass in the NEM.
• Convert biomass to biogass and run a gas engine. Very flexible. Turn up and down in a few
minutes. Similar to liquid fuels Bagasse used to fuel a boiler to run a steam turbine- less
flexible. Typically run all day to prevent restart.
• Generally operated on the basis of the underlying plant i.e. waste processing of sugar cane
from sugar mills

Question

Compare relative operating flexibility of different electricity fuel generation technologies and explain
how each can be advantageous in helping meet demand. Include characteristics such as ramp rates,
start times, fuel costs, minimum operating capacity, and reliability.

Answer

Coal has historically been the workhorse of the electricity industry because the plants are relatively
cheap to build, and there is an abundance of readily available and cheap fuel. Coals drawback is that
it is a steam turbine generator and therefore relatively slow to ramp up and down. It is also timely
and costly to start and stop( Start times. Cold=12hrs, Warm=between cold and hot, Hot=1hr)
Coal’s main contribution to meeting demand is supplying base load due to its predictability and
low fuel cost. Coal also has a relatively high minimum operating level (30%-50%) and so bids zero
or negative prices to remain on. Coal is relatively operationally inflexible.

Gas plant has more expensive fuel costs than coal (approx 4x), and has much faster ramp rates and
start times (10-20 mins, w/ fast start less than 5min). They also have lower minimum operating
levels (20%-30%).This makes them useful as a peaking plant to meet rapid increases in demand.
If gas is already committed then it can also provide faster ramp down rates than coal for sudden
losses of load. Gas is used for some base load but more for higher daytime demand.

PV and wind have incredibly fast ramp rates relative, essentially zero start/stop times or costs,
generate using non-storable, discontinuous resources, zero fuel costs, and no minimum operating
level. PV can be advantageous during peak demand events on hot days. Typically if it’s a very
hot day there is plenty of sunlight to power PV and it can offset additional air conditioning loads.
Hydro also has almost no start time and fast ramp rates. This is due to it being a non-thermal
turbine generator. It also has the advantage of zero fuel costs and being dispatchable in the short
term. Hydro generates strategically when prices are high so that it doesn’t waste much of it
reserves. It also doesn’t require electricity to start and thus can be used to black start the grid.

2
Liquid fuels like diesel and kerosene that run combustion engine generators are the fastest plants in
the NEM, both for starting and ramping (seconds to minutes). They are used for meeting critical
peak loads. This is because they are by far some of the most expensive operating plants in the
NEM (and these demand peaks cause high enough prices for liquid fuel to be economic) and they
can start and ramp with little notice. These fast dispatch plants operate infrequently, but when
they do it helps prevent blackouts in the network.

Nuclear is even more operationally inflexible than coal, and for the same reasons as coal, it is
typically used to meet base load demand. Optimal operating strategy for nuclear is run for 2
years; stop for refuelling and maintenance; start up; and repeat. This is possible as there is
generally about 2-3 years of fuel in the fuel rods. Very low fuel cost. Heat from nuclear reaction
runs steam turbine. The danger is that altering the rate of reaction too much increases the risk of
malfunction. The Chernobyl disaster was caused by minimum operating level checks.

Question

Define SRMC. Explain variable operating costs in relation to time scale. Given a gas power
plant with fuel cost 10 $/GJ, sent out thermal efficiency of the gas plant at 21 GJ/MWh, variable
operating and maintenance cost of 14 $/MWh (sent out), start-up cost of $200, and auxiliary power
of 4.5% generated. Calculate SRMCSO and SRMCG .

Answer

Short run marginal cost is the cost to a producer (generally measured in monetary terms) incurred
by producing one extra unit in the short run. Alternatively, it is the change in short run total costs
for an extremely small change in output. Short run costs are non-fixed costs. Long run includes
all costs to the producer.

Variable operating costs are the increase in costs incurred to the producer when the output of
the business (generator) changes. In this course, it is generally taken to mean the increase in fuel
and maintenance costs of the plant and are given in ($/MWh). The time horizon dictates what is
included in the variable costs. The longer the time horizon, the more costs get included into the
variable cost (and the less costs are counted as fixed. Eventually the time horizon is long enough
that all costs are considered variable. This is no longer short run but long run.

First, let’s convert efficiency to a percentage.

GJ GJ 109 × M W h
21 = 21 × 6 = 5.83333
MWh M W h 10 × 602 × GJ

This is bigger than 1, so it’s actually the inverse of efficiency. (Note the units are energy in divided
by energy out.)

So the efficiency is 1/5.8333 = 17.1%

3
 SRM CSO = 10 $/GJ ÷ 0.171 + 14 $/MWh
= 5.8333 $/MWh + 14 $/MWh
= 19.8333 $/MWh
SRM CG = 10 $/GJ + 14 $/MWh × 0.171
MWh
= 10 $/MWh × + 2.394 $/MWh
GJ
106 × 602 J
= 10 $/MWh × + 2.394 $/MWh
109 J
= 10 $/MWh × 3.6 + 2.394 $/MWh
= 36 $/MWh + 2.394 $/MWh
= 38.394 $/MWh

Question

Describe the average household demand profile of the NEM during a typical summer week and a
typical winter week and briefly explain any differences throughout the week and between seasons.

Answer

Typical weekday summer house load profile has a small morning peak that drops slightly then
continues to increase, plateauing about 8pm and then dropping quickly as the night continues
down to early morning levels.

Typical winter household load profile has a large, spikey peak in the morning, dropping considerably
during the day followed by a big increase in demand in the early evening which gradually decreases
to a low around 4am, ready for the next morning’s peak

Thermal comfort typically dictates the shape of a household’s demand profile. In summer people
run air-conditioners in the early afternoon and into the evening to cool their house. Cooling is not
generally required during the morning as the house is cooler from the night. On the other hand,
in winter households use a larger amount of heating (space, food and water) during the morning
before work/school and in the evening when people return home for the night to maintain thermal
comfort. These habits form the peaks and troughs of a typical weekday load profile for summer
and winter. Typically, weekends have a lower peak demand and so are smoother load profiles than
weekdays. This is due to the relative lack of time constraints that most people face during the
week; namely working 9am-5pm and running the majority of their appliances to accommodate
their working lifestyle. Weekends do not have this constraint and thus appliance demand can be
spread out at the user’s leisure.

Loads are low when there is little activity in the household. Inactive periods are generally between
10pm and 5am and during the middle of the day.

Question

How and why might a plant’s actual operation differ from its theoretical operation? Provide at
least 2 examples.

4
Answer

Plant operation is a little more flexible than standard operating model. For example, it’s not
uncommon for generators to run at 105% of their rated capacity for a short period of time. This
typically occurs during times high prices in the spot market where generators are trying to maximise
their dispatch. Presumably, the returns from running above the rated capacity of a plant for a
short period of time outweigh the costs due to factors such as maintenance and reduced plant
lifetime.

Another way in which a plants theoretical operation can be different to its actual is through the
ability of generators to rebid their ramp rates in the market. Ramp rate rebidding is allowed
because it can physically change depending on a particular plants situation. However, a plant can
rebid its ramp rate so that it doesn’t reflect the operational ability of the plant. This can keep
them dispatched longer if they are anticipating to be turned down/constrained off, or dispatched
slower if they expect to be constrained on. Generally there are monetary gains to be made by
doing so. Similarly a plant can withdraw its capacity (economic withholding) from the market if it
does not wish to be dispatched, or if it wants to attempt to tighten the market in the short term.

Generators that rebid ramp rates or withhold capacity must provide a reason to AEMO as part of
the good faith agreement. These reasons can be fairly opaque and don’t generally get scrutinised
by AEMO, allowing a generator to act away from its theoretical operation.

Topic 2 Week 4: Economic and Market Dispatch

Question

Calculate the incremental variable cost of an OCGT plant with the following technical parameters.
The OCGT plant operates using coal seam methane with a cost of 2.94 $/GJ. The fixed operating
costs are 14307 $/MW/year and variable operating and maintenance costs of 8.73 $/MWh sent out.
The plant has an emissions factor of 55 kgCO2 /GJ and a thermal efficiency of 30%. There is no
carbon price.

Answer

V OC = Fuel cost + V OM
= 2.94 $/GJ ÷ 0.3 + 8.73 $/MWh
109 · M W h
= 2.94 $/GJ$/M W h × ÷ 0.3 + 8.73 $/MWh
106 × 602 · GJ
= 2.94 $/MWh × 3.6 ÷ 0.3 + 8.73 $/MWh
= 44.01 $/MWh

Question

What is operational decision making in the electricity industry as a centralised engineering opti-
misation problem?

5
Answer

The electricity industry decision making as a centralised problem is when one body can make the
decisions for the EI given an inventory of generation, network and demand side electricity equip-
ment, all having technical parameters, operating costs, industry benefits and operating constraints
wrapped up in uncertainty. The centralised body has the ability to control all the above. Using
what they are given they calculate a strategy to maximise IBOT.

This is typically a stochastic nonlinear dynamic optimisation problem for operating decisions in
generation, networks and demand side.

Question

What are the objectives and challenges of electricity industry restructuring?

Answer

In an ideal world, centralised decision making would be the way to conduct the electricity in-
dustry, but in reality this only leads to low innovation, poorly behaved participants and a non
consumer based industry. The objectives are to enhance economic efficiency, social accountability
an environmental sustainability. The challenges are:

• shared essential nature of network services

• collective (demand and supply) responsibility for availability and quality of supply
• exposing decision makers to the associated risk

Australia has restructured using markets as a framework for decision making, governed by the
theory of supply and demand, allocating resources through price mechanisms. A market makes
the assumption that electricity is a commodity that can be traded.

Question

What is operational decision making in the EI as a decentralised engineering optimisation problem?

Answer

As a decentralised optimisation problem, decision making in the NEM becomes a problem involving
knowing all existing gens, network and demand side elec equipment, their technical parameters,
operating constraint, operating costs and derived “energy service” benefits of demand. Included in
this decision making are all external costs and benefits associated with operation of these options
and the uncertainty of the future. The decentralised decision makers have the ability to control
their own generation, network or end use.

This is achieved by establishing markets that maximise social benefit. Market participants un-
dertake operational decisions that contribute to maximising societal welfare over operational time
horizon.

6
Question

the electricity is one large problem with many operational challenges, how is this solved in the
NEM? Include the engineering optimisation and the economic optimisation tools used and their
practical implementation.

Answer

The NEM solves this by breaking down the problem by time.

• Seconds to minutes

– Engineering optimisation uses automatic generation control

– Economic optimisation uses ancillary services markets
– Practical implementation hybrid engineering and commercial arrangements
• Minutes to hours

– eng: Economic dispatch of operating units

– eco: spot market
– prac: wholesale market subject to network and plant constraints
• Hours to days

– eng: unit commitment of available units

– eco: forward markets
– Prac: short term forward markets aided by centralised system adequacy assessment
(PASA)
• Weeks to months

– scheduling maintenance
– Derivatives
– derivative markets and PASA
• Years

– Not so much an operational problem but is planning for engineering and derivatives for
eco, practical is policy development.

Topic 3 Week 5: Market dispatch in the NEM

Question

Determine the market price, and the dispatch and profits of each generator in the following single
node scenario.

7
Answer

Gen 1 = 700 MW
V1 (P1 ) = 20P1 Demand 1 = 800 MW
B1 (D1 ) = 300D1
Gen 2 = 400 MW
V2 (P2 ) = 40P2
Demand 2 = 200 MW
Gen 3 = 200 MW B2 (D2 ) = 80D2
V3 (P3 ) = 100P3

Price ($/MWh)
300

100
80

40
20
0 Demand (MW)
700 800 1000 1100 1300

Generator Dispatch (MW) Profit ($/h)

1 700 1400
2 300 0
3 0 0

Question

Determine the market price, and the dispatch and profits of each generator in the following single
node scenario. The link is lossless but constrained.

8
Answer

Gen 1 = 700 MW
V1 (P1 ) = 20P1 Demand 1 = 800 MW
B1 (D1 ) = 300D1
Gen 2 = 400 MW
V2 (P2 ) = 40P2 200 MW limit
Demand 2 = 200 MW
Gen 3 = 200 MW B2 (D2 ) = 80D2
V3 (P3 ) = 100P3

Price ($/MWh)
300

100
80

40
20
0 Demand (MW)
700 800 900 1000 1100

Generator Dispatch (MW) Profit ($/h)

1 700 42000
2 200 8000
3 0 0

Question

Determine the market price, and the dispatch and profits of each generator in the following single
node scenario. The link is lossy but unconstrained.

9
Answer

Gen 1 = 700 MW
V1 (P1 ) = 20P1 Demand 1 = 800 MW
B1 (D1 ) = 300D1
Gen 2 = 400 MW
V2 (P2 ) = 40P2 30% loss
Demand 2 = 200 MW
Gen 3 = 200 MW B2 (D2 ) = 80D2
V3 (P3 ) = 100P3

Price ($/MWh)
300

100
80
57

20
0 Demand (MW)
700 800 980 1180

Generator Dispatch (MW) Profit ($/h)

1 700 42000
400 at generator
2 8280
280 at bus
3 0 0

Question

Determine the market price at each node, the dispatch and profits of each generator, and the power
through each transmission line in the following 3-node scenario. Each line has equal reactance and
no losses.

10
 G1: 1000 MW G2: 1000 MW
$20/MWh $40/MWh

C3: 900 MW

Answer

G1: 1000 MW G2: 1000 MW

$20/MWh $40/MWh
900 MW Dispatch 0 MW Dispatch
Price: 20 $/MWh Price: 20 $/MWh
300 MW

600 MW 300 MW

Price: 20 $/MWh
C3: 900 MW

Generator Dispatch (MW) Profit ($/h)

1 900 0
2 0 0

Question

Determine the market price at each node, the dispatch and profits of each generator, and the power
through each transmission line in the following 3-node scenario. Each line has equal reactance and
no losses.

11
 G1: 1000 MW G2: 1000 MW
$20/MWh $40/MWh

100 MW limit

C3: 900 MW

Answer

G1: 1000 MW G2: 1000 MW

$20/MWh $40/MWh
600 MW Dispatch 300 MW Dispatch
Price: 20 $/MWh Price: 40 $/MWh
100 MW limit
100 MW

600 MW 300 MW

Price: 30 $/MWh
C3: 900 MW

Generator Dispatch (MW) Profit ($/h)

1 600 0
2 300 -3000

Question

Describe some of the key features of the NEM. How do generators participate? Who else partici-
pates?

Answer
• Covers Qld, NSW, Vic, SA and Tas
• Multi region gross wholesale electricity spot market with dynamic intra-regional loss factors

12
 • Separate node, and therefore price for each state (region)
• Generators bid price/quantity pairs a day ahead, with an opportunity to re-bid up to 1
minute before each five minute time period
• Most load bid in at market ceiling price of 13,500$/MWh
• Hybrid 5/30 minute dispatch to maximise benefits-costs – i.e. bids are made for 5 minute
intervals which determines the spot price for that interval, but the actual price paid to
generators is determined every half hour (generators are paid the spot price of the last 5
minute interval for all generation in the previous half hour)
• 8 Frequency Control Ancillary Services for maintaining supply/demand balance over time
periods less than 5 min
• There is no capacity market or equivalent – participants must determine unit commitment
through an energy spot market bidding strategy
• AEMO provides all the necessary information for generators to determine their unit commit-
ment strategy, including weather forecasting, demand forecasting and historical data
• Compulsory participants: all dispatchable generators and links > 30 MW (unless intermit-
tent) and network service providers and retailers

Question

The spot price is determined in such a way as to maximise the value of spot market trading.
Identify 5 factors which affect the formation of the spot price (there are 11 in total)

Answer
• Dispatch offers and bids, and ancillary service offers
• Constraints due to availability and commitment
• Non-scheduled load requirements in each region
• Power system security requirements
• Intra-regional network constraints and losses
• Inter-regional network constraints and losses
• Constraints consistent with registered bid and offer data
• Current levels of dispatch
• Ancillary service requirements
• Pro-rata loading of tied bid and/or offer data
• Minimising the impact of a direction or reserve contract dispatch

Question

Explain what market power is, and how it can be exercised by a large generator (a diagram may
be useful). What can be done to mitigate market power?

13
Answer
• Ability of one or more participants to influence market price – can either be through a
monopoly power or collusion
• Some degree of market power inevitable, e.g. due to constrained conditions or genuine com-
petitive advantage

• A large generator can offer some portion of its capacity at a very high price, and the rest at a
much lower price (to ensure dispatch). If a sufficient portion of supply is withheld or offered
at a high enough price, this can cause the market clearing price to increase dramatically,
ensuring high profits for the capacity that has been offered in at a low bid. (see lecture notes
for diagrams)

• Can be mitigated through market structure such as increasing number of generators, avoiding
base load and peaking plant in one portfolio and enhancing distributed resources, or through
market design by reducing benefits/increasing risks of gaming, facilitating entry by generators
and demand response, preventing artificial supply constraints and implementing efficient
retail markets

Topic 4 FCAS and large scale failures

Question

What are the 4 key technical characteristics that Ancillary services (AS) control and for what
purposes? How is the NEM’s AS structured?

Answer

The 4 key technical characteristics are frequency, voltage, network loading and system restart
processes. With its 3 services (FCAS for frequency, NCAS for voltage and network loading, SRAS
for system restart), AEMO has to manage the power with 3 goals: safety, security and reliability.

FCAS is a market-based AS, it is jointly monitored with the spot market and constitutes of approx.
75% of the AS costs. NCAS and SRAS are non-market AS and they usually need very specific
types of generators (Hydro for SRAS for example)

Question

What are the types of frequency changes that FCAS has to face and what are the technical solutions
provided to bring back the frequency to 50 Hz?

Answer

When there is a generator failure/outage, there is a lot more pressure put from the loads to the
supply side, so the frequency drops out. On the contrary, when there is a transmission line failure
or when there is a load outage, the generators put a lot more pressure on the demand side, so the
frequency increases. FCAS has to face two types of events in the NEM:

14
Common small adjustments called regulation (maintenance of a generator or a transmission line. . . ):
these events are neutralised thanks to inertia, load relief and to centralized engineering decisions
made by AGC (Automatic Generation Control; linked with AEMO).

Uncommon large events called contingency (generator or load outage, transmission line failure. . . ).
FCAS has 6 services based on the time response: 6 seconds, 60 seconds and 5 minutes, for raise
and lower. For lower, the 6 seconds and 60 seconds services are usually operated by load shedding
(or governor response), 5 minutes services re-establish the frequency thanks to rapid generators.
“Layering” contingency services allows units with diverse properties to participate in those markets
that suit their technical and economic characteristics.

Topic 5 Unit commitment, financial instruments and schedul-

ing production costing

Question

What is unit commitment and how is the optimal path chosen?

Answer

• Unit commitment is the determination of which generators must make themselves available
for dispatch and at what times/dates.
• Constraints on ramp-rate and the cost of turning on/off means that intertemporal links need
to be considered in order to minimise the total cost of meeting demand over time. Operation
and control techniques need to think ahead to compensate for the lag in adjusting which
generators are on and at what volume. It is for this reason that unit commitment is it’s own
process that precedes and provides boundaries for economic dispatch.
• The options for determining the least cost solution centrally are:
– Exhaustive search by enumeration, where the number of possibilities to be compared
equals (2N − 1)M where N is the number of generation units and M is number of
commitment periods.
– Dynamic programming (less computing power needed)
– Shut-down priority list
– Mixed integer-linear programming

• What’s problematic with centralised unit commitment that the optimization is performed
on the assumption that the constraint information (about max/min levels, ramp rates, turn
on/off times/costs) is accurate, when in reality generators game the system to maximise
profits

• In the NEM however, unit commitment is left to the generators to perform, using detailed
demand and price forecasts provided by AEMO. The underlying idea is that by providing
generators with as much information as possible they will make themselves available in the
lowest cost combination.l

15
Question

What are some of the common financial instruments used in the electricity industry and how do
they reduce risk for market participants?

Answer

Common instruments in the EI include:

• Two-sided contracts for difference (CFD)

– Generators sell to retailers
– Allows retailers/generators to fully hedge their position to protect against future price
risk, which requires an estimation of both future spot price and cumulative volume
– Generators aren’t protected against outage risk
• Call option or cap (one-sided CFD)

– A type of insurance where the buyer pays an option fee such that if the spot price is
above the strike price they are compensated by the seller.
– Might be bought by a consumer with inflexible demand to protect them against exposure
to high spot prices, or by an unreliable base load generator to protect them against the
risk of being unavailable to generate during high price events.
– Might be sold by reliable high cost peaking plants to ensure a stable income regardless
of the frequency of high price events (ie their income is the insurance fee, and if they
have to pay out they are compensated by generating and selling energy at the high spot
price). The option fee is set in order to cover fixed operating costs, or allow for the
fulfilment of minimum take or pay gas contracts, or guarantee return on investment.
• Put option or floor (one-sided CFD)

• A collar is a combination of a call option and a put option to put effective boundaries on the
extremity of the costs the buyer can be exposed to.
• Only relationship between financial instruments and the physical product is through the spot
price because the quantity can be anything. The disincentive to contract for much higher
volumes than you can control is a financial risk.

Question

What is production costing, what is it’s function and what are the key challenges?

Answer
• Production costing is the simulation of future operation of the existing power system for a
specified time period with a focus on dispatch and spot price
• Production costing’s function:
– Allows for the estimation of future operating costs & reliability,
– Informs market participants so they can make better choices

16
 – Simulates system operation with improvements/expansions of the power system, as well
as failures in the power system

• Key challenges:
– Determining supply and demand-side operating costs (start-up/shut-down, fixed & vari-
able costs, maintenance costs, network losses, cost of unmet demand)
– Characterising supply and demand-side uncertainty (generator/network/large load out-
ages, demand forecast uncertainty ie due to weather dependence)
– Simulating performance

Question

How does the primary energy constraint impact operation for hydro and peak plants?

Answer

• Primary energy (or access to it) is constrained for all forms of electricity generation, and in
many cases the constraint is such that the fuel supply could not allow the plant to operate
at maximum output indefinitely.
• For hydro this is the case because of the finite capacity of reservoirs and the uncontrollable
nature of rainfall. The impact upon operation is that generators must ration the primary
energy source so that they can generate when prices are highest. This involves considering
inter-temporal links in decision making, with storage level as a state variable. Dynamic
programming is useful optimisation tool.
• The benefits of this are twofold: the hydro generator ensures that it is maximising revenue
and is displacing high cost alternatives which would otherwise be dispatched during high
price events.
• With hydro, the level of water in the reservoir dictates the minimum spot price at which they
will generate, for example if the reservoir is overflowing then they will generate at a very low
spot price but if the reservoir is nearly empty then they will generate closer to the VOLL.
• For peaking plants, the primary energy constraint means they must decide tactically the size
and timing of fuel purchases in order to maximise the value they get for using the fuel.
• Maintenance outage scheduling also needs to be performed in order to ensure that the plant
is available during high price events.

Topic 6 Renewable energy and distributed resources

Question

Describe the consequential challenges resulting from high RE penetrations

Answer
• High RE penetration leads to increased variability in net demand in electricity grid

17
 • This brings about frequency control problems, which increases the need for inertia control,
regulation and additional ramping requirements
• In order to deal with the varying dispatch nature of renewables, thermal plants require higher,
more frequent levels of cycling operations (involving starts/stops)
• Thermal plants have to adjust their outputs more quickly and frequently, which is challenging
towards their technical limits

Question

What are some changes that need to be implemented in order to accommodate high RE penetration
in the grid?

Answer
• Accommodating high RE penetration in future electricity industry begins from the generation
planning stage
– To ensure that future power systems can accommodate high renewables
– Future generation portfolios must have adequate flexibility to cope with sudden and
frequent changes in generation
• Changes in existing market mechanisms may also be required to ensure revenue sufficiency
for market generators
– This may involve capacity payment mechanisms or increasing/removing the market price
cap
– Operational practices usually involve
– Increasing the number of interconnections, increasing the flexibility of thermal plants
and considering demand side options
– Operational limits involving a minimum synchronous generation are also implemented
• Most generation planning and investment modelling techniques largely ignore the operational
characteristics and impacts of renewables
– Long-term planning models usually don’t consider inter-temporal operational aspects
– Optimal portfolios obtained from LT planning models may not be operationally or
economically viable

Question

List all the distributed energy technical options within the EI that actively participate in EI
decision-making

Answer
• Renewable energy sources, including solar thermal, PV, smaller-scale wind and biomass
• Small-scale fossil fuel generation, combined heat and power CHP plants powered with engines,
gas turbines or fuel cells

18
 • Direct energy storage; chemical ‘battery’ technologies, flywheels

• Electrical end-use equipment (They actively respond to changing conditions) i.e. ‘smart
buildings’ that control heating & cooling to exploit their inherent thermal energy storage
• End-use energy efficiency and smart devices

Question

Discuss the challenges and opportunities for distributed energy in the EI

Answer

Challenges

• It is challenging to establish appropriate EI arrangements that allow DE to receive both

• Energy & network values, given that they are highly location-specific, time-varying and have
a contingent value of energy and network flows (spot & future prices affected by decisions
made, which impact future decisions)

• Environmental values, which may require regulations like the ETS, MRET and feed-in tariffs
to internalize environmental & social externalities
• In addition, there are many issues surrounding wholesale & retail market design, network
regulation and policy frameworks for restructured industries

• It is difficult to maintain technology and participant neutrality, given the varying technical
& economic characteristics, location and ownership profiles by end-users
• Retail markets where DE resides are the ‘unfinished’ business of many EI restructuring
processes

• Existing EI involves the intersection of regulated network and competitive supply/demand

options that is invariably complex and imperfect
• Present jurisdictions and policy developments are not yet seriously addressing environmental,
energy security and issues of the wider social externalities of energy markets

Opportunities

• Some highly cost-effective alternatives to centralized supply and associated network options

• Environmental benefits from use of RE resources of highly efficient fossil-fuel use

• DE allows for greater end-user engagement in achieving desired energy services

Question

Discuss the potential future opportunities in demand-side participation for Distributed Energy
Options

19
Answer

Many key development opportunities for the EI lie alongside the demand-side in current ‘retail’
markets-energy efficiency, demand-side participation, distributed generation

• This however, requires that end-users be ‘ready, willing and able’ to act
• This includes support for Energy Service Companies that will be focused on the delivery of
energy services

• More economically efficient retail + network ‘prices’ can help

• A coherent and comprehensive policy, market and regulatory framework should be developed
to facilitate instead of discriminate against new societally valuable options
• Network decision making could utilize an Integrated Resource Planning approach rather
than current arrangements, allowing for -Greater transparency, stakeholder participation
and separation of market power monopolies in the EI

Topic 7 Assignment 2

Question

Is market power always a bad thing?

Answer

Market power is very often a bad thing. It distorts the markets, and can lead to more expensive
generators being dispatched over cheaper ones. This is non-optimal and reduces IBOT. Market
power of incumbents can also be a barrier to entry of new market participants.

However, in the NEM, some market power is a good thing. For example, the generators on the
margin do not make a profit. Therefore they need to exercise some market power to make a profit,
so that they can pay off their capital costs and so on.

Question

The following table lists the minimum operating levels, capacity, incremental cost and emission
intensity of 3 generators. Assume floor price is -1000$/MWh, and the wind farm is generating at
50% of its full capacity. Draw the system-lambda curve ($/MWh vs MW) for this network.

Generator Capacity (MW) Minimum Incremental Emissions Intensity

Operating Cost (tCO2e /MWh)
Level (MW) ($/MWh)
Coal 400 200 20 1
Gas 300 100 60 0.5
Wind 400 0 10 0

20
Answer

Price ($/MWh)

Gas
60

Coal
20
Wind
10
Demand
300 500 700 900 (MW)

Coal & Gas

Minimum
Operating
Level
-1000
-100

Question

How much would a carbon price need to be to reduce the emissions of a network with the generators
listed above?

Answer

Wind is already dispatched before coal and gas. Gas is cleaner than coal, so we need the cost of
coal ($/MWh) to be greater than the cost of gas.

Let x be the carbon price ($/tCO2e)

coal cost > gas cost

x × 1 tCO2 e/MWh + 20 $/MWh > x × 0.5 tCO2 e/MWh + 60 $/MWh
x × 0.5 tCO2 e/MWh > 40 $/MWh
x > 80 $/tCO2 e

Therefore the Carbon Price must be at least 80 $/tCO2 e.

21
Question

For the generators listed above, for a demand of 800 MW, how could the coal plant operator adjust
their bids to exercise market power and increase their profit? Assume the wind and gas plants are
bidding non-strategically, and the coal plant operator has full knowledge of their costs. How much
more profit does each of the 3 plants make, compared to non strategic bidding? Does total system
surplus change?

Answer

The coal plant can change their bid price to the ceiling price. Assuming the ceiling price is
13500 $/MWh. They will get dispatched at a lower volume (from 400 MW to 300 MW), but the
higher spot price will more than make up for that. Their new profit is

300 MW × (13500 $/MWh − 20 $/MWh) = 4044000 $/h

The spot price for this level of demand without strategic bidding was 60 $/MWh. So their previous
profit was
400 MW × (60 $/MWh − 20 $/MWh) = 16000 $/h
. Therefore the increase in their profit is

4044000 $/h − 16000 $/h = 4028000 $/h

The wind plant will still get dispatched at 200 MW. Their profit is increased by

(13500 $/MWh − 60 $/MWh) × 200 MW = 2688000 $/h

The gas plant will get a dispatch of 300 MW. So their profit will be

300 MW × (13500 $/MWh − 60 $/MWh) = 4032000 $/h

Their profit when there was no strategic bidding was 0 $/h, since they were on the margin. There-
fore their increase in profit is 4032000 $/h.

With strategic bidding, the more expensive gas plant is getting 100 MW more dispatch, and the
coal plant is getting 100 MW less dispatch than with non-strategic bidding. Therefore the total
system surplus is reduced, because the new dispatch is sub-optimal. The total system surplus is
reduced by
100 MW × (60 $/MWh − 20 $/MWh) = 4000 $/h.

Question

If all customers in the NEM were given prices reflective of the spot market (e.g. time of use pricing,
critical peak pricing, spot market price etc) instead of a flat tariff, would that increase, decrease,
or make no difference to the amount of market power the generators have?

Answer

• The motivation behind giving customers cost-reflective pricing is to reduce demand during
periods of network stress.

22
 • One way for generators to exercise market power is to withdraw volume, to increase network
stress, to drive prices up. Alternatively, they can increase the price of a bid, which can drive
the price up (if that price-pair quantity becomes marginal).
• When generators exercise market power, they (generally) make the price go up. Without cost-
reflective pricing, the load would not change much as a consequence. With cost reflective
pricing, the load would drop off as the price goes up. Consequently, generators would get
less dispatch than otherwise, and hence their increase in profit would be lessened. Therefore,
widespread cost-reflective pricing reduces their market power.
• The amount by which demand drops off as the spot price rises would be variable. Therefore
this adds an extra factor of uncertainty for the generators. Increased uncertainty reduces the
generator’s ability to exercise market power.

Topic 8 Assignment 3

Question

Use forward dynamic programming to solve unit commitment for the following:

• 2 generators, with the costs and constraints outlined in the following table
• Value of lost load at 13500 $/MWh

• Demand in hourly windows, as specified in the table below.

• Initial state of gen 1 on and gen 2 off
• Ignore shutdown costs

Generator Minimum Startup No load cost Variable Operating Capacity

Operating Costs ($) ($/h) Cost ($/MWh) (MW)
Level (MW)
1 50 5000 1000 30 200
2 100 10000 2000 20 500

Time Interval 1 2 3 4
Demand (MW) 150 200 250 150

Answer
Note: I have probably made a mistake in this question. It’s super easy to do. If
you spot a mistake, email vicepresident@elsoc.net.

There are 3 different levels of demand, and 4 possible states. We can assume that having both
generators off will not be the best solution for, so we have 9 states we need to solve for. So we
need to solve economic dispatch for 9 cases.

23
 Demand (MW)
State
150 200 250
Gen 1 only (A) 1000 + 30 × 100 = 1000 + 30 × 150 = 1000+30×150+50×
4000 5500 13500 = 680500
Gen 2 only (B) 2000+50×20 = 3000 2000 + 100 × 20 = 2000 + 150 × 20 =
4000 5000
Both on (C) 1000 + 2000 = 3000 1000 + 2000 + 50 × 1000 + 2000 + 100 ×
20 = 4000 20 = 5000

Now we work out the cost to transition between each state

Transition Cost ($) To

Gen 1 only (A) Gen 2 only (B) Both on (C)
From Gen 1 only (A) 0 10000 10000
Gen 2 only (B) 5000 0 5000
Both on (C) 0 0 0

Now to put it together

24
Time 1 2 3 4
period
Demand 150 200 250 150
(MW)
Gen 1 only 4000 Min of: ∞ (close enough) Min of:
(A)
• A: 4000 + 5500 = • B: 22000+5000+
9500 4000 = 31000
• B: 13000+5500+ • C: 23000 +
5000 = 23500 4000 = 27000
• C: 13000 +
5500 = 18500

Gen 2 only 10000 + Min of: Min of: Min of:

(B) 3000 =
13000 • A: 4000+10000+ • A: 9500+10000+ • B: 22000 +
4000 = 18000 5000 = 24500 3000 = 25000
• B: 3500 + • B: 17000 + • C: 23000 +
23000 = 26500 5000 = 22000 3000 = 26000
• C: 13000 + • C: 17000 +
4000 = 17000 5000 = 22000

Both on 10000 + Min of: Min of: Min of:

(C) 3000 =
13000 • A: 4000+10000+ • A: 9500+10000+ • B: 22000+5000+
4000 = 18000 5000 = 24500 3000 = 30000
• B: 3500 + 5000 + • B: 17000+5000+ • C: 23000 +
13000 = 21500 5000 = 27000 3000 = 26000

• C: 13000 + • C: 17000 +
4000 = 17000 5000 = 23000

Optimal B B or C B B
State

25