OZ-ENERGY-ANALYSIS.ORG - open science for the new millennium
THE STORIES | DATA | ANALYSIS | MODELS | LITERATURE | DISCUSSIONS
Status: [Aug 2010] working draft; static
Terms used:
NEM = National Energy Market
AEMO = Australian Energy Market Operator
The job of the electricity system is to provide electricity as needed to meet demand, with this process corresponding to the areas of Supply, Transmission and Demand. For most Australians (SA, VIC, TAS, NSW, QLD) the management of the electricity system has been integrated into a National Electricity Market (NEM). This page provides an overview of how the NEM operates, and includes some background on the Transmission and Distribution Networks, and the way in which electricity is moved through these from generators to loads.
The loads on the system change constantly, and the system as a whole must be managed and adjusted to provide the supply commensurately. Three quite distinct time scales are involved in this constant balancing. Briefly, for a changing system load, in the first few seconds any extra electricity needed is drawn from the mechanical inertia of generators; in the range between around 6 seconds and 5 minutes an "Ancillary Services" market responds to system feedbacks; then in 5 minute chunks the "Dispatch" is controlled by an electricity market (which prices wholesale electricity every thirty minutes).
The introduction of inherently variable power sources (e.g. Wind Turbines), reduces control over system stability. The AEMO have sophisticated mechanisms to forecast the wind resource, and complex forecasting extends to the system as a whole. A model of the entire system is linearly optimised every 5 minutes in order to manage the dispatch (to match supply to demand). All of the above aspects will be described in what follows.
The Australian electricity system runs at 50 Hz (the reader is assumed to have a basic understanding of AC electricity). Generators, of what ever sort, must run in a synchronised way, otherwise the system can be analogous with two car batteries side by side and wired -ve to +ve at one end, and +ve to -ve at the other. As for these DC batteries, a system with multiple connected AC generators would quickly burn out or blow up without synchronisation. So, when a generator comes online it must first be synchronised with the system; once matching waveforms are established the generator can be switched in. Once part of the system (and as a basic understanding) a generator will stay synchronised as the system overall undergoes variations in the AC frequency. In the case of the NEM the frequency is regulated by the AEMO to stay within a band 49.85 to 50.15 Hz.
In the case of a large steam turbine, such as will be found in a large coal-fired power station, it is the physical speed of turbine rotation that defines the frequency of the electricity. Sudden or unanticipated changes in the system load have a direct effect on system frequency as electrical energy is exchanged with the kinetic energy (the rotational momentum) of the large turbines. So, for example, if the electrical load increases, with no change in the steam powering the turbine, then the turbine will slow down (and the electrical frequency will decrease). The energy to make the extra electricity is being taken from the kinetic energy of the turbine. Conversely, if the electrical load reduces (with no change in the steam powering the turbine), the turbine will speed up as the extra energy is absorbed into the kinetic energy of the turbine. It is the kinetic energy of the large turbines that provide the first line in managing system stability.
Distinguishing between the Transmission and Distribution Networks is useful and real. Bulk power is carried at high voltage (up to 500,000 V) by the Transmission System from generators to substations around the country; while the Distribution Network carries electricity at lower voltages (ultimately 240 V AC for residential consumers) from substations to individual users. As a rule of thumb about 10% of electricity is lost in the networks, and thus the overall generation supply needs to be about 10% more than total demand. Note that at OzEA we are primarily concerned with the Transmission Network.
A brief note on transformers and voltage stepping: The great advantage of AC electricity, and the principal reason for its use, is the ease with which the voltage can be stepped up or down with transformers. This is important because high voltages are required to efficiently move bulk electricity through wires over distances, and these high voltages can be progressively stepped down into lesser branches of the transmission network, and through into the distribution network. The reason for needing high voltages can be appreciated using two pieces of elementary physics: V=IR and P=IV, where I is current, V is voltage, R is Resistance, and P is Power. Because the transmitted Power is a simple product of the Voltage and the Current, the same amount of power (ignoring the losses in the wires) can be transmitted with double the voltage and half the current (for example). Consider: a current I in the wire gives the relationship I = ΔV / R (i.e. V=IR) where ΔV is the voltage drop across the wire; now, the power lost in the wire is P = ΔV.I = I2.R (by substitution), thus showing that resistive losses in transmission grow with the square of the current, and thus that high-voltage and low-current minimises resistive transmission losses.
The Distribution Networks, mostly utilising voltages up to 130 kV, are owned at the state level by commercial, or corporatised government, entities. In SA the distribution network is owned and operated by Electranet (and others in other states). In many parts of the country the distribution networks are struggling under peak loads, and need either to be upgraded, or need for Demand Management (via Smart Grids and Smart Metering) to better manage these peaks, and thus postpone or even eliminate the need for further 'poles and wires'.
The Transmission Network, mostly utilising voltages between 130 kV and 500 kV, constitutes major infrastructure connecting Generators to the substations that then distribute power to the various Residential, Commercial and Industrial customers. Any substantial new sources of generation require transmission infrastructure. More information about the owners of Australia's $10 billion worth of transmission networks can be found on the Grid Australia website (http://www.gridaustralia.com.au/).
The Interconnectors are that part of the transmission network linking the individual NEM states, including (for example) the Basslink between Tasmania and the mainland. The [various] interconnectors have carrying capacity ranging from 170 MW to 3000 MW, and include both high voltage AC and DC transmission. The inclusion of large amounts of renewable power into the NEM system may require significant upgrades and / or expansion of the interconnectors.
Predominantly, supply is managed and dispatched to meet demand (while to a much lesser extent demand is managed to meet supply). As introduced, there are three distinct levels at which this is achieved: first (as above in the section on System Frequency), the physical inertia of large turbines flattens out immediate fluctuations in the load, and essentially converts these into fluctuations in system frequency. There needs to be enough physical inertia in the system to maintain system stability. It is the legislated role of the AEMO to maintain the system frequency in the range 49.85 to 50.15 Hz for at least 99% of the time (AEMO 2009), while ensuring "shortfall energy" (i.e. failure to maintain continuity of supply) is no greater than 0.002% of total annual energy provided in any given region (AEMO 2009), which they do through the remaining two levels of system control.
The Ancillary Services Market manages supply in the 6 sec to 5 minutes range, principally by working to regulate frequency. After system inertia has managed the first few seconds of a normal but unanticipated variation in load, AEMO calls various fast-ramp-rate generators into action, and these respond by immediately changing their power output. This is the Ancillary Services Market - it is dedicated to supplying or removing power on call, and is generally constituted by open cycle gas turbines (OCGT) and/or Hydropower. While OCGT are inefficient compared with a combined cycle gas turbines (CCGT), they are able to provide the fast response that is needed. Ancillary services have a range of +-250MW in the NEM, and support system control in the up to 5 minute range.
[Oct 2010] The AEMO also manage the system to hold up under foreseeable disruptions or breakdowns (such as the sudden loss of a large generator), and here Ancillary Services play a key role (see comment #5 and surrounding for now).
Dispatch and Semi-Dispatch: After this time frame comes the Dispatch regulating method. This is the guts of the NEM operation. In this case generators are called to dispatch power from the AEMO. The AEMO is given a contract from the generators at 12.30pm the day before, outlining in 5 minute intervals how much power they will provide and at what price they will provide it. This is called generator bidding and allows the AEMO to call those generators to provide up to what they said they would when required. There is some flexibility for the generators here, they can change their offer up to 5 min before being called to dispatch, only the price they initially offered remains the same.
There is one other mode of power contribution, Semi Dispatch, which is relevant to intermittent supply at a nameplate rating of greater than 30 MW. It was introduced in 2005 and implemented in 2008 as a means of increasing system stability control. Prior to the semi dispatch regulation, the wind farms would have all of their produced power contributed to the NEM; this made it increasingly difficult to regulate supply and hence control system stability.
Forecasting is very important in the process of matching supply with demand as it allows accurate preparation prior to the real time event. There are complex models and systems used by the AEMO in order to forecast demand, generator bids, and wind output. Essentially, if AEMO know what power will be required and what generation capacity will be available at what price, they can ensure system stability.
Demand forecasting occurs through neural network forecasting models. Significant work has been done at Monash University, with the main input being temperature and calendar effects to predict the peak summer and winter demand in SA for a ten year period (Hyndman and Fan 2009).
Supply Forecasting relies on complex game theory to predict generator behaviour. As the generators are all bidding to optimise their profits, the dispatch bidding is essentially a share market. Generators also have complex contracts between themselves that mitigate the risk of the price volatility.
Wind Forecasting is essential for large-scale wind farm electricity contribution. AWEFS (Australian Wind Energy Forecasting System) is the main body of work contributing to this (AEMO 2010).
Dispatch is determined by AMEO (every five minutes) based on solving a linear system of some 15,000 constraint equations. This approach may not always be optimal, however, the constrained linear optimisation algorithm is guaranteed to provide a solution for the dispatch. The system constraints are vast and complex, including the transmission limitations and generator bidding prices. In this way the system progresses through time in five minute chunks, and supplies the electricity that we all demand.
Maintaining system stability is a primary function of the NEM / AEMO; or, more bluntly, supply must be constantly managed in order to meet demand (including in the event of sudden system failures). Electricity is supplied through large networks of transmission and distribution infrastructure, including interstate links, with many voltage changes along the way.
System inertia manages fluctuations in overall demand for the first few seconds, with ancillary services regulating supply in the up-to 5 minute range. The market-based system around dispatch is the mechanism by which the NEM matches supply to demand on an ongoing basis.
The AEMO have sophisticated systems that forecast the generators dispatch behaviour based on complex game theory; they use neural networks to forecast the demand behaviour; and have a strong focus on wind forecasting through atmospheric changes. In total this large system of interactions and inputs act to provide control over system stability and to reliably match supply to demand. This system will continue to develop and evolve, particularly with the challenges provided by the continuing integration of renewable energy sources into the NEM.
References:
AEMO 2009, An Introduction to Australia's National Electricity Market, p. 10
AEMO 2009, Power system frequency and time deviation monitoring in the NEM, table A1, p23.
Hyndman and Fan 2009, Forecasting long-term peak half-hourly electricity demand for South Australia, Monash University.
AEMO 2010, Australian Wind Energy Forecasting System (AWEFS) overview.
| 2 |
Neil Howes |
| 3 |
Arthur |
| 4 |
Francis |
| 5 |
Arthur |
| 6 |
Francis |
| 7 |
DV82XL |
as & fc - August 2010
Post Comment: