Anatomy of an energy crisis - a pictorial guide part 1

Who could forget the energy “crises” that affected electricity supply across south-eastern Australia last year.  The Conversation

What energy crisis?

Firstly the Tasmanian crisis, following the  Basslink outage in December 2015.  With hydro storage dams at record lows following a drought on the back of aggressive storage withdrawals during the carbon tax years, Tasmania enforced drastic measures to ensure supply.   Thankfully, flooding winter rains, together with the eventual restoration of Basslink in June helped resuscitate the apple isle’s energy supply. Tasmania’s hydro storages now stand at around 40% of full capacity, more than double at the same time last year.

Tasmanian hydropower storage capacity shows a strong seasonal trend, filling in winter rains, and drawing down during the summer and early autumn. Exchanges with Victoria via Basslink help provide security of supply, that was compromised by the outage in December 2015, when storages were already dangerously low on the back of the  drought conditions in 2015 and aggressive draw down of storages during the Carbon-tax years to capitilise on the higher mainland spot prices.

July saw the first of the sequence of crises in South Australia that followed from, and were in many eyes, attributable to the closure  of its last coal-fired power plant at Port Augusta in May of 2016.

With gas prices at record highs, and South Australia effectively isolated from Victoria due to upgrades on the main interconnector into Victoria, spot prices sky rocketed, culminating on a cold, windless winter day on July 7th. Energy consumers that had not contracted supply were at the whims of traders. Prices averaged over $1400/MWhour for the day and around $520/MWhour for the week, almost 800% above the average for that time of year.

Graphical summary of electrical power generation,  demand,  spot prices in, and exchange between, each of the five regions comprising the National Electricty Market.  The period shown is the week of July 3rd- 9th, 2016, during the first South Australian energy crisis. Over the week,  interconnector flows from Victoria into South Australia were restricted to an average of 225 MW, or about 40% of full capacity due to upgrade works. On July 7th, at the height of the crisis,  the flow was limited to 166 MW. VWP = volume weighted price in $/MWhour. TOTAL.DEMAND = regional demand in MW. DISPATCH.GEN = regional generation in MW. NETINTERCHANE = net exports (positive) or imports (negative) in MW.

All that was superseded by the events of September, when extreme winds played havoc with the South Australian transmission system, toppling transmission  lines in the mid north. Poorly understood default control settings automatically disconnected  wind farms, leading to the  interconnector tripping and a  state-wide black out. Unanticipated problems in restarting the system exacerbated the pain.

Finally,  failure of a transmission line in south-west Victoria on December 1 lead to a power loss at the aluminium smelter in Portland. The damage to “frozen” pot lines has jeopardised the smelter’s ongoing viability. As the state’s largest energy consumer and the one of the biggest regional employers, the political fallout is intense.

After the NEM’s “annus horibilus”

With 2016 very much the National Electricity Market’s (NEM) “annus horibilus”, pundits awaited the summer of 2017 with bated breath. The combination of high gas prices, frighteningly intense summer heat, a fragile and ageing energy supply system, and increasing concerns about market rules,  the scene was set for “interesting times”. Whatever was to transpire  it was always going to be inflamed by political point-scoring -  the one commodity that seems rarely  in short supply.

And so it would prove to be, even in the northern states of Queensland and New South Wales that had hither-too largely escaped the wrath of Electryone.

The summer of 2017 has seen extraordinary rises in spot prices beset the NEM,  particularly in New South Wales and Queensland.  Further blackouts in South Australia, and  market interventions to avert them in New South Wales, have done little to assuage concern if our electrical power system is fit for purpose.   So far, 2017 Queensland spot prices have been around 400% above the historical average for this time of year.

Graphical summary of NEM operations  for the period 1st January - 11th February 2017.

With the summer far from finished, our politicians remain hard at it, pointing fingers and apportioning blame,  doing almost anything and everything but that which is in most short supply - namely, embracing bipartisanship. A glimmer of hope is to be found in comments from Chief Scientist Alan Finkel, who has been charged to lead a review of the security of our National Electricity   Market.

What is the NEM?

To provide some guide to what is happening to the NEM, and why, I have compiled a  few pictures that illustrate elements its  basic anatomy. This is designed as background. In following posts in this series I will focus on the details of recent events that have so heightened the political heat.

The  NEM comprises five interconnected regional jurisdictions -  one for each state along the eastern seaboard and South Australia. For each region, the market operator AEMO runs a  5-minute interval, energy-only, dispatch ‘pool’, or spot market. The objective is to balance supply with demand in a way that minimises cost, based on the bids submitted by generators. It is a complicated process.  Settlement prices are aggregated at half hourly intervals, and determined as the average of bid prices of the last offer needed to meet the demand for the dispatch interval.  

Pictorial of the generation structure on the NEM, as of early 2017. The top half shows the five regions comprising the NEM, the bottom half the power as generated and dispatched by fuel type progressing from fossil on the left through to  renewables of the right. For the period shown (1/1/2017-11/2/2017) black coal contributed 55.6% of supply (at a capacity factor of 68%), brown coal 22.7% (cf=79%), natural gas 11%  (cf = 24%), hydro 5.4% (cf = 14%) and Wind 4.6% (cf=29%) Units are in MW. Note that gas is the only fuel source common to all regions, but its contribution varies significantly from over 50% in South Australia, to just a few percent in Victoria. It is important to note that the small-scale distributed solar PV on domestic rooftops is not dispatched to the NEM and so is not included in this graphic. The installed capacity of around 5 GW of small scale solar PV contributed about an average of 700-800 MW across 2016.

With the focus of the dispatch ‘pool’ being least cost electricity supply, AEMO also operates several ancillary markets  to ensure the  requirements for safe grid operation are met. This includes the provision of reserve supply and frequency control normally sourced from synchronous generators such as large coal plants.  

AEMO also has regulatory powers to intervene in the market by demanding  generation be made available in cases when the total bid capacity is insufficient.  When demand exceeds total capacity, or if the available  capacity cannot be made available in a timely fashion,  AEMO can authorise load-shedding, effecting a re-balancing of demand to meet the available generation capacity.

Normally, large electricity consumers will contract power supply via the contract market, rather than directly through the spot market. This insures consumers against the potential for extreme price volatility allowed on the spot market, that can see prices range from between -$1000 and $14,000/MWhour.  For comparison, the standard domestic retail tariff is about $250/MWhour  or $0.25/kWhour.

The bid strategies of power plants reflect differences in their cost structures and performance characteristics. For example, fuel costs for brown coal generators are very low, but they are best operated at constant load.  In contrast gas plants are generally much more rampable, but much higher cost. In Victoria, as a consequence gas is used almost exclusively to meet peaks in demand as illustrated in the three graphics below.

Dispatch in Victoria for the period 8/2/2017-10/2/2017, coloured by fuel source. Also shown is the Victorian demand (brown line), available generation bid into the market (top black line), and net exports as negative (bottom black line)
Brown coal power generation  in Victoria for the period 8/2/2017-10/2/2017, coloured by power station.
Natural gas  generation in Victoria for the period 8/2/2017-10/2/2017, coloured by power station.

Typically a large base-load generator, such as a brown coal plant, will bid much of its capacity into the spot market at their short run cost, to ensure a slice of the action. In contrast peaking power plants will bid at price well above marginal cost, anticipating that they will required only very occasionally.   Forward contracts of various kinds help insure revenue streams for base load generators against spot prices below their long-term cost of production, and for peaking plants being available when needed.

Renewables such as wind  dispatch at the whims of the  weather, and because of negligible  short run marginal costs, bid their output at very low prices. As a price taker, wind generation tends to drive spot prices lower, impacting the viability of other generators.  As shown below, and to be discussed in more detail in a following posts, the recent events in South Australian dispatch highlights the challenges in the market when wind power output correlates poorly with demand.

Dispatch in South Australia for the period 8/2/2017 through 10/2/2017, coloured by fuel source. Also shown is the South Australian demand (brown line), available generation bid into the market (top black line), and net imports  (bottom white line). Black outs  on the 8th February occurred when local dispatch curve hit the available generation. At that time here was no more capacity ready to be dispatched, so AEMO instigated load-shedding. (Note that not all capacity in South Australia was bid into the market at this time.)

Finally, rooftop PV is not dispatched onto the grid, but rather is “revealed” to the market as a reduction in demand.

Why are spot prices rising?

In theory, the  spot market is designed to encourage a competition that ensures prices provide generators with a revenue stream that is linked to their long run marginal cost of production.  If prices do depart, competitive market principles should ensure system re-balancing  either through investment in new generation or the withdrawal of old. Of course, competition needs to be provided  by an adequate  diversity in ownership.

And so shifts in the spot prices, signalled via the contract markets, are designed to reflect the balance of demand and supply. The years 2009-2014 were characterised by persistent reductions in demand across the NEM, in part due to growing penetration of solar PV. At the same time, the addition  of new wind farms to meet Renewable Energy Target  contributed  to a growing oversupply in the market, reflected in very subdued spot prices. For example from 2010-2014, Victorian spot prices averaged about $35/MWhour, after factoring out the carbon tax. While that price is above the  cost of production for existing Victorian brown coal generators, it would be well nigh impossible to obtain financing for any new large scale generation at prices less than about 2-3 times that.

Since 2014,  demand has risen in Queensland due in part to the commissioning of new LNG gas processing facilities at Curtis Island. Reductions in generation capacity in  Victoria and South Australia due to closure and/or mothballing of several fossil plants (Anglesea in Victoria and Northern and Pelican Point in South Australia), has significantly tightened the supply-demand balance. Consequently, spot prices are on the  rise across the NEM.

Why do spot prices vary  between regions?

Spot prices  averaged about $60/MWhour across last year, but vary somewhat by region  and by season.

As shown in  diagrams above the  make-up of generation in each of the five regions varies considerably, leading to different cost structures. Similarly differences in demand profiles lead naturally to differences in generation fleet. Finally there are  differences in  market competition.

With limited interconnection capacity, along with differences in regional demand and generation portfolios, occasionally lead to large separation in spot market prices.  In times of very high demand during summer heat waves and  winder cold snaps, or in times when supply is constrained by infrastructure (power plant or transmission) outages or fuel supply/cost issues, spot prices can be extremely volatile.

Annual variations in spot prices for the period 2nd January through 11th February, for  each of the four mainland regions. Red numbers shows the average for the years prior to 2017.

Historically, South Australia has had the  highest  prices and Victoria the lowest. This reflects the much higher proportion of gas in the generation mix, its  larger proportional  daily/seasonal cycle  between minimum and maximum demand and, arguably, competition issues. As illustrated below, peak demand in South  Australia is over 250% higher than the median, compared to around 150% in  Queensland. A greater relative proportion of peaking generation capacity means higher average spot prices. Competition is a particular issue in South Australia, since the closure of the Northern Power Station, as it is in Queensland.

Annual demand in South Australia and Queensland, in MW in top panel, and as percentage of median demand in bottom panel. Note the recent rise in demand in QLD due in large part to the recent commissioning of LNG plants. The bottom panel highlights the high variability in demand in SA which sees maximum demand reaching up to 2.4 times the median. In comparison QLD peaks  only reach about 1.5 times median.   The boxes show 25-75 percent quartile ranges with notch at the median. Outliers more than 1.5 times IQR are shown by dots.

How well suited is our market?

It is important to realise that while the physical characteristics  of any power system are governed by the laws of physics,  the market itself is  a construct - just one of many ways of matching  supply and demand. In particular as an energy-only ‘pool’ , there are questions about how well our NEM is suited to meeting the  need of providing a cost effective, secure and environmentally acceptable energy supply. In particular, there is very little incentive for demand side management. Moreover, the power system does not operate in isolation, and needs to be considered with other policy settings in the gas and water markets as well as climate policy. In the following posts in this series I intend to address some of these issues with examples drawn from our recent experience on the NEM.

Mike Sandiford, Chair of Geology & Redmond Barry Distinguished Professor, University of Melbourne

This article was originally published on The Conversation. Read the original article.

More Information

Mike Sandiford

mikes@unimelb.edu.au

Receive the latest in energy news from across the University Subscribe here