Home | About Us | eSubscribe | Links |        
   CSIRO  |  SOLVE  | Issue 3 - May 05  
ARTICLE
ENERGY:
Power Surge
By GRAEME O'NEILL

IN THE COMPLEX ELECTRICITY TRADING MARKET, HUMAN BRAINPOWER ISN'T ENOUGH.

One of the reasons that industry in eastern Australia was traditionally able to briefly shut down for the Melbourne Cup and then restart without causing a catastrophic overload in the electricity grid was the Snowy Mountains Hydro-Electric Scheme. Its water-driven turbines could adjust in seconds as electrical equipment was turned off then on again after the race.

On Cup Day the Snowy provided base-load power, backing off as race time neared then, as workers returned to their desks and machinery, kicking in again until the slower-responding thermal power stations took over.

It was a delicate procedure by power station operators, but almost quaint compared to the way this rapid response capacity is used today in the now deregulated, volatile electricity market in which demand and supply and prices fluctuate constantly. The Snowy now uses its rapid-response capacity to sell 'insurance'; increasing supply (when called on by 'policy holders') to dampen demand-driven price spikes when they threaten the value of forward contracts with electricity retailers and consumers.

Picture: Robin JareauxIt is a fast-moving, highly complex trade, and will intensify even further next year when Tasmanian hydro-electricity also becomes available to the National Electricity Market (NEM) through the Basslink cable. NEM was created in 1998 by the four eastern mainland states and the ACT.

Generator companies bid on a daily basis to supply electricity wholesalers. They must adapt to volatile prices and patterns of demand that fluctuate with the time of the day or week, the season and shifts in the weather. To maintain competitive advantage they must also monitor the bidding activity of their competitors.

The corporate players in this restructured power business confront complex decisions daily. Experience helps, but making sound economic decisions in the face of uncertainty demands something more than human brainpower.

Enter NEMSIM - the National Electricity Market Simulator, a powerful computer model being developed by scientists within CSIRO's Energy Transformed Flagship. NEMSIM is integral to the Flagship's mission to develop new, efficient and low-emission energy technologies and systems for Australia.

NEMSIM's project leader, Dr George Grozev, says NEM is well designed and its performance ranks with the world's best. But with the world confronting the threat of global warming, Australia, like other nations, is under pressure to meet rising demand for energy, even as it implements conservation measures to reduce greenhouse gas (GHG) emissions from fossil fuel burning.

The electricity industry is responsible for around 35 per cent of Australia's GHG emissions. Australia is heavily reliant on coal - 85 per cent of Australia's electrical energy comes from coal-fired power stations, and they are a major source of GHG emissions.

"Electricity is not like most other goods," Dr Grozev points out. "It can't be stored efficiently, and supply and demand must be matched instantaneously."

Demand often fluctuates widely from minute to minute, according to the time of day or sudden weather changes. For example, a cold front passing over a capital city after a heatwave can send temperatures plunging by 10 degrees or more in 20 minutes, causing a rapid, massive drop in energy consumption as hundreds of thousands of air-conditioners switch off.

Coal-fired power stations cannot adjust their output at this timescale. Gas turbines and hydro-electric generators provide flexibility, with their ability to be brought rapidly on line, or turned up or down, but they account for only 15 per cent of Australia's total generation capacity.

Dr Grozev says that as new or improved technologies become available, NEMSIM will help operators to explore the commercial consequences of their decisions, and their effects on GHG emissions, before they actually make them. They will be guided on whether, when, where and how much to invest in new infrastructure - new generating units and transmission lines - to meet projected energy demands.

A powerful feature of NEMSIM is its use of computer-simulated 'agents' to represent power-generating companies, network service providers, retail companies and a market regulator.

    At a Glance
  • MAINTAINING a constant flow of power in the deregulated, highly volatile electricity market is a fast-moving and highly complex trading environment
  • SUPPLY and demand must be matched almost instantaneously
  • NEMSIM, a new computer modelling system developed under CSIRO's Energy Transformed Flagship, allows the major players to make better decisions, creating a more efficient and stable industry
"Trying to model human behaviour and decision making in a state of uncertainty is usually very difficult to do," Dr Grozev says. "Some other modelling tools are very useful, but not strong in this respect.

"We expect our system will have some advantages, by allowing for more complicated decision making. For example, it will allow agents to use different strategies for bidding, learn from their own experience or that of more successful agents and change their particular strategy. Look-ahead functions allow a user to identify the best bid among a set of alternative options."

While 'intelligent'agents cannot control the behaviour of rivals, they can adapt their own. To simulate such behaviour, the model draws upon advances in gaming theory, and economic modelling approaches pioneered 50 years ago by brilliant American mathematician Professor John F. Nash, of 'A Beautiful Mind' fame.

Dr Grozev says each corporate agent acts independently within the system, according to its own goals, to increase its profitability within the constraints and rules applying to its particular role. It may pursue a strategy of maximising its market share at a relatively small profit margin, or choose to diversify into alternative power-generation sources such as gas-powered turbines, or 'renewables' like solar, wind, biomass, tidal or thermal power.

Agents can make routine daily decisions, such as using particular bidding tactics for the next day, negotiate medium-term hedging contracts to reduce exposure to high price volatility or make long-term investments in new infrastructure.

NEMSIM provides a platform in which the population of simulated agents interacts according to market rules and the physical constraints of Australia's power grid and associated infrastructure.

Dr Grozev says the agents'individual and collective behaviours co-evolve from the bottom up, and market outcomes emerge spontaneously from the model's internal dynamics.

NEMSIM is still under development, and new features are still being added, but it is approaching a stage where players in the power industry will be able to play with it, ask 'what if' questions based on various assumptions, and receive answers in the form of likely outcomes or scenarios.

Developing NEMSIM involves a collective, multidisciplinary approach. Under the umbrella of the Energy Transformed Flagship, CSIRO's Manufacturing and Infrastructure Technology, Energy Technology and Atmospheric Research divisions are partners in the project - the last is contributing its expertise in modelling weather and climate.

"The main thing is that, with current modelling techniques and advances in computing power, we can perform very sophisticated simulations that were previously unthinkable," Dr Grozev said. "In the short term, NEMSIM can solve problems to help power companies improve their bidding in the very dynamic daily energy market.

"Every generator company makes its bids for the next day, for each of its generating units, but then they can re-bid, or change the quantities they are bidding for, according to changing circumstances or their competitors' behaviour. Some re-bid 10 to 20 times a day."

Dr Grozev says NEMSIM is also a powerful modelling tool for power-generation companies to explore investment in extra generation capacity to meet future demand.

Timing is also critical - like Rome, a major power station is not built in a day. Several years ago, California experienced a series of 'brownouts' during heatwave conditions, because power generators had postponed investment in new capacity. Network service providers can use NEMSIM to make decisions to extend or upgrade parts of the network to accommodate growing demand, or changing patterns of demand.

High-voltage power lines 'leak' energy to the environment as heat. The further the electricity is transmitted, the higher the energy losses. Increasing generation capacity can be an expensive option if the extra capacity is required only for short periods. The alternatives are to improve the network, or move to distributed generation.

And NEMSIM is not just for power utilities. Dr Grozev says it could be used in applications such as GHG emissions trading schemes such as one already in operation in Europe. NEMSIM contains a GHG emissions calculator that could be extended to allow government agencies to simulate different trading rules and allocation plans.

CSIRO is now seeking to commercialise NEMSIM, and to enhance the software.

 

 

For further information contact:
CSIRO Enquiries
Email: Solve@csiro.au      Web: www.csiro.au
Tel: 1300 363 400       International: +61 3 9545 2176

Back to main
 

Solve Issue 3 - May 05
IN THIS ISSUE

 

 

Last Updated: May 5, 2005
© 2005 CSIRO Australia. For use of CSIRO material contact solve@csiro.au
 
Use of this website and content is subject to our Legal Notice and Privacy Statement.
Please contact us for assistance, or to provide feedback or comments.