AN OPTIMAL PORTFOLIO OF NEW POWER
GENERATION TECHNOLOGIES:
AN ILLUSTRATION FOR SOUTH AUSTRALIA
Anthony D Owen and Ntasha Berry
International Energy Policy Institute
UCL Australia, Adelaide
tony.owen@ucl.ac.uk
Presentation to the 32nd USAEE/IAEE North American Conference
Anchorage, Alaska, 25-26 July 2013
Australia’s National Electricity Market:
an energy-only market
• Gross pool: obligatory market
• Price bids by generators every 5 minutes, averaged
over half-hour trading intervals in five regional spot
markets
• Typical bid structure
• Some MW @ very low or –ve $/MWh: up to minimum load
• Some MW at SRMC values: up to contracted position
• Rest at blue sky values
• All dispatched power attracts the same marginal bid
price, irrespective of the above
• Floor price = -$1000/MWh
• Ceiling price = $13,100/MWh
The Cumulative Price Threshold
If the sum of the trading interval spot prices over a
rolling seven day period total or exceed the Cumulative
Price Threshold (CPT), then spot prices are capped at
the administrative price cap of $300/MWh. The CPT is
currently $197,100, and is subject to annual indexation.
Characteristics of the SA market
• Tail-end of longest connected grid in the world –
5000 km;
• Electricity demand very “peaky” – peak load (in
excess of 2,600 MW) occurs around 2% of year;
• Load ranges between 1,200 MW and 2,400 MW
90% of the year;
• Installed capacity (nameplate) 4,000 MW;
• Installed capacity of wind – 1,203 MW;
• Dominant generator – gas thermal (1,280 MW).
SA generation mix
Gas
Wind
Coal
Capacity
55%
24%
15%
Generation
50%
26%
24%
On particular days, wind has accounted for up to 65 per
cent of total generation in the state, and up to 86 per cent
of generation for a trading interval. However, wind
generation is generally lower at times of peak demand; on
average, it contributes to less than nine per cent of supply
at any given time during summer.
Number of negative trading intervals
in the NEM
2009-2010
86
2010-2011
208
2011-2012
274
Given the RECs wind receives, it can bid negative
to ensure dispatch. Coal and gas may also bid
negative to avoid costs associated with shutting
down or operating inefficiently.
Electricity market intervention
Current government policy in the power sector is
primarily focused on policies to support the
development and deployment of non-nuclear lowcarbon technologies to reduce their costs and thus
reduce the long-term costs of de-carbonising the
sector. However, the current NEM market design may
make low-carbon investments riskier than continued
investment in fossil fuel technologies. Thus, even with a
carbon price, investment in low-carbon technologies
may be discouraged.
Carbon pricing
The marginal generator (that sets the wholesale
price) is generally a fossil fuel generator. Thus,
marginal bid pricing should include the carbon
price. This should benefit low carbon technologies.
However, the risks attached to recovery of fixed
costs of new generation assets will vary
considerably depending on the capital intensity of
the different technologies. For large up-front cost
low carbon technologies (such as geothermal,
concentrated solar thermal, off-shore wind, CCS,
etc.) these risks may more than offset the cost
minimisation criteria for investment choice.
Investment in power generation
It is intended that price “signals” should encourage
investment in new capacity, but long-term investment
decisions appear to conflict with the short-term (i.e.
instantaneous) nature of the market and may impede the
effectiveness of climate policies. Uncertainty on future
electricity prices is compounded by other risks:
1. Long-term uncertainty on carbon price;
2. Unclear competitive nature of low carbon technologies;
3. Long lead times for high up-front cost low-carbon
technologies: the chain of innovations is too long, too
complex and too diverse;
4. High levels of political and regulatory risks; and
5. Uncertain availability of finance.
Wholesale market without wind
Price ($/MWh)
Demand
WP
Infra-marginal
rents
Capacity (MW)
Baseload technologies
Intermediate technologies
Peaking technologies
Wholesale market with deployment of
wind
Overview of investment issues
Classification of investment opportunities
Investment opportunities in the NEM may be
classified as:
capacity-driven: short-fall in reserve margin
reflected in high prices during peak demand,
usually requires investment in OCGT;
energy-driven: determined by longer term
(average) price trends; and
policy-driven: arise from Government policies,
generally to encourage renewables.
Uncertainties in the SA electricity market
• Demand (SA): mining expansion/industrial &
manufacturing contraction + PV penetration.
• Fuel price (International): domestic gas prices
are relatively low by international standards,
but will rise as LNG netback prices kick-in.
• Carbon policy (National): high degree of
uncertainty with linking to EU ETS, but could
be abandoned altogether!
21 possible scenarios to reflect interactions
Least cost generation expansion modelling
This study developed a least-cost generation expansion model to
determine a set of long-term generation options from a set of possible
technologies which minimised the cost over time, taking into account
future demand, carbon price, and gas price uncertainties. For this
application, the long-term planning simulation module of a
commercially available software, Plexos, was used.
The model emulates the process performed by the dispatch system in
the NEM, which schedules generators that minimise the dispatch cost,
subject to the criteria that the reliability and security of the system is
maintained. The model executes half-hourly dispatch using security
constrained optimum generation dispatch with each generator bidding
at its short-run marginal cost.
Model constraints
• An energy balance constraint, which ensures that the total
electricity demanded at a given time is equal to the un-served
energy at that time plus the sum total of the electricity dispatched
by all the generation units at that time;
• A feasible energy dispatch constraint, which ensures that the level
of electricity dispatched by a given generator is not greater than the
maximum generating capacity of that generator;
• A feasible build constraint, that ensures that the built capacity (in
MW) for a given new entrant technology does not exceed the value
of maximum built capacity (as set in the model) for that technology
in a year (or over a given planning horizon); and
• A reliability standard constraint, which ensures sufficient generation
capacity to meet demand.
Model inputs
The model considered the cost parameters of the existing generators
in terms of their fixed and variable operations and maintenance costs,
combined with their technical constraints: ramp-up and ramp-down
rates, thermal efficiency, combustion emission factor, fugitive emission
factor, minimum stable generation level, forced outage and planned
outage rate, auxiliary load requirements and marginal loss factor over
the planning horizon. Variation of nameplate capacity of generators
due to seasonal factors was based upon past experience.
The model expanded the generation capacity within the region based
on the inter-region transfer capability. The interconnector transfer
capacity limits the transfer of energy between Victoria and South
Australia during any load block and affects the reserve sharing
capability between the two regions.
Cost data for new entrant generators
CCGT
Nameplate capacity (MW)
Capital cost ($/kW)
Variable O&M cost
($/MWh)
Fixed O&M cost
($/kW/year)
Discount rate
Economic Life (years)
Carbon price ($/tonne)
Fuel cost ($/GJ)
Thermal efficiency (%)
Capacity factor (%)
LCOE ($/MWh)
OCGT
Geotherm
EGS
Geotherm
HSA
Solar
Wind
374
1062
160
785
10
10,600
20
7,000
10
8,308
100
2,530
4
9.3
-
-
-
-
10
2.5
170
200
60
40
6.81
30
23
6.42
49.5
83
75
6.81
30
23
6.42
33.5
10
188
6.81
30
83
138
6.81
30
83
104
6.81
30
30
272
6.81
20
35
90
Results: Energy, capacity and policy-driven
investment opportunity in SA
The optimum energy, capacity, and policy-driven
investment opportunities in SA in terms of capacity,
technology and timing of additional generation over the
next 20 years are given in the next slide. As a result of the
closure of coal fired power plants (under the CEF), the
reserve margin fell significantly and short-term
investment in OCGTs was required. However, the results
are dominated by the LRET constraint driving investment
in substantial amounts of wind energy. Solar thermal and
geothermal HSA technologies were also stimulated by the
LRET, with initial build commencing in 2016 and 2019 and
reaching maximum regional build by 2020.
Energy, capacity and policy-driven
investment opportunity in SA
Conclusions I
As might be expected, the LRET scheme, “peaky”
demand, the high level of (intermittent) wind capacity,
and volatile gas prices favoured wind and OCGT
technology.
Fuel price risk is the main source of concern for
investments in CCGT technology, although generators
with an upstream gas position have a natural hedge
against this risk. In addition, the less capital intensive
OCGTs can have their cycle “closed” when deemed
appropriate. The “peaky” nature of demand in SA makes
this an ideal strategy for meeting any future increase in
base load demand.
Conclusions II
The modelling also indicated that geothermal
HSA technology was to be preferred on a cost
basis to CCGT for base load, largely due to high
gas price expectations and low operating
capacity for the latter. However, this result is
predicated upon geothermal becoming
financially viable by 2020.