Models, forecasts, scenario projections … they’re all viewed with a certain degree of scepticism. And, for the most part, rightly so – we’ll never know the future with complete certainty.
But what modelling does provide is a way to prepare for an uncertain future, by identifying risks and opportunities and evaluating alternative options when times get tough.
Think about the weather. Should I hang my washing out today? Do I need to pack a coat or sunscreen for my holiday next week?
We can be reasonably confident planning our daily and weekly activities based on weather forecasts. More general projections of a wet or dry year are also fairly reliable and very important for the agricultural industry.
But what about forecasting Australia’s future energy needs for, say, 2050? Despite the uncertainty in such a task, every government and major energy company in the world is involved in forecasting the future of energy for one simple reason: infrastructure.
Big investments need reliable projections
Most energy infrastructure has a 25-50 year lifespan. Our largest power stations are built with a 40 year lifespan in mind, but that can be extended to 50 years with the help of refurbishments.
One example is the Munmorah Power Station on the New South Wales Central Coast. It’s still operating despite being commissioned 42 years ago.
Given the multi-billion-dollar cost of such infrastructure, government and energy companies need to know what needs to be in place as society changes and populations shift.
And, when you have to make payments on a multi-billion-dollar loan for the next 30 years, you want to know that what you’re building will be relevant to energy use demands of the next 25-50 years.
A small investment in futures research can narrow the range of possible outcomes when someone is making an energy investment. It not only makes sense, but is often a requirement when convincing financiers to back a project.
What’s this energy modelling all about?
Energy modelling is a challenging multidisciplinary task that brings together many elements. It has to look at societal changes, economic factors, environmental impacts and technology development.
In order to work on projects that will have an impact in the future, we need to know what society, the economy and our environment will need in 20-50 years time.
Energy forecasting and modelling is also invaluable in making informed policy and investment decisions.
When industry or government wants to understand the prospects of a new energy project, they model the whole energy market, all of the competing technologies and fuels, and the changing needs of consumers and industries using that energy source.
Consider a proposal for a large-scale solar plant. As energy modellers, here are some of the questions we’d need to be able to answer:
- What is the future wholesale price of electricity?
- How will the cost of solar energy and its competitors improve over time?
- What’s the best location for the project, given the location of the existing grid connection points and the variation in sunlight across the country?
- How does a carbon price and other policies change the above factors?
By answering these questions we can then guide and inform investment decisions made by government and industry.
It’s all connected
The value of modelling lies in the accuracy of projections. As modellers here at CSIRO, there are a number of paths we pursue to continuously improve the rigour of our modelling tools.
We are developing new ways of integrating different forecasting models from different sectors. For example, we no longer view forecasting from agriculture, land use and energy as separate items.
We understand how interconnected these sectors are and it’s vital to consider the whole in our increasingly interconnected world.
Yet each sector has its own separate modelling communities, languages and preferred approaches. To connect them all we are continuously developing new mathematical techniques and systems.
The importance of being open
Transparency is vital when we are using models to argue for alternative policies. In a perfect world everyone would conduct modelling on an objective basis, but in reality we know that some modelling is biased by the group that commissioned the modelling.
Interest groups may model only selected scenarios that support their agenda. A stakeholder seeking to lobby against a proposed policy change will have an incentive to select a more pessimistic set of scenarios describing the impact of the proposed policy, rather than modelling the full range of possibilities.
Similarly, stakeholders in support of the same policy change might only examine optimistic scenarios.
When we are conducting modelling for our own purposes we seek to do so objectively. We develop a balanced scenario set and provide details of any assumptions we’ve made so our objectivity can be tested.
When we are commissioned to conduct modelling, the commissioning party chooses the scenarios. We try to make those scenario assumptions transparent so the reader can make up their own mind about the outcomes of the modelling.
Preparing for an uncertain future
Modelling, forecasting and scenario projections are part of everyday life decisions, from deciding whether or not to take an umbrella to work, to an energy company investing billions of dollars in new technology.
It is not crystal ball work – it is a scientific approach, combining a multitude of variables, data and industry characteristics to understand likely outcomes.
It is critical for our society and our energy future to make decisions based on the outcomes data modelling provides. Without it we would find it much more difficult to evaluate the risks, opportunities and potential impacts our decisions might have on our nation’s environment, society and economy.
In fact, guesswork is about the best we could do.
CSIRO’s modelling work includes the following, recent reports: