+44 (0) 1865 952700 |
Press enquiries
  • This field is for validation purposes and should be left unchanged.


Home/Time for a smarter approach to power price forecasting

Time for a smarter approach to power price forecasting

Earlier this month, Polish utility Polenergia made headlines after obtaining close to $30 million in debt financing for a 38 MW wind farm – without the certainty provided either by government support or a long-term power purchase agreement (PPA). This is part of a broader trend across Europe as a combination of dwindling state subsidies and falling costs are driving an increasing number of wind and solar projects to seek financing on a purely merchant basis.

In this context, it is increasingly important for developers, financiers and equity investors to think hard about how best to quantify merchant risk. Historically, the approach was fairly simple: renewable capture prices tracked the baseload price, which in turn closely tracked the clean gas price. Today, things are less straightforward. The more wind and solar on the system, the more renewable capture prices diverge from the baseload price as cannibalisation increases. This new market complexity, combined with a growing merchant exposure, demands a fresh approach to power price forecasting.

How low is “low”?

How have market forecasters responded to this increasing complexity? In short, not especially well. For years, the received wisdom has been that price risk could be assessed via “high” and “low” case sensitivities around a central forecast. But how low is “low”? How do forecasters decide which underlying assumptions to flex – and by how much? And what is the probability that actual prices will under- or overperform the low case?

At Aurora, we thought we could do a bit better. That’s why we developed our P90 capture price methodology. Over the past 18 months, from Poland in the east to Portugal in the west, we have used this approach to help assets secure financing—and so far, demand shows no signs of abating. The Polenergia project cited above is just one example.

Playing with probabilities

Volume risk is already well understood – there are good wind years and bad wind years – and using P10/P90 weather years has become the industry standard. Price risk, by contrast, is trickier to quantify. At a high level, our approach follows four key steps:

  1. Identify individual price risks, from commodity prices to cannibalisation drivers and the level of system flexibility
  2. Determine realistic high/low bounds for each individual risk and assess the impact in each case on baseload and renewables capture prices
  3. Analyse the relationship between different risks and the likelihood of them occurring at the same time
  4. Model the impact of all possible combinations of risks to describe an overall distribution of power price outcomes, and use this to define a low case which equates to a P90 of this distribution

Of course, there’s only so much statistical analysis that can be done around discrete policy choices like how much nuclear or interconnection the government is likely to support. No methodology is completely free from subjective judgment, no matter how much that judgment is based on sound reasoning. What matters most is being as robust and transparent as possible such that all input assumptions are readily justifiable. Above all else, this is what clients tell us they most value about our approach.

It’s perhaps little wonder that others are beginning to follow our lead. Certainly other market forecast providers appear to moving away from a generic “low” case * and coming around to the approach Aurora has been using for much of the last two years.

Looking ahead – time to get specific

At Aurora, being innovative is in our DNA. From building all our own models to integrate capacity, balancing and ancillary markets, to coding cutting edge battery dispatch algorithms and modelling the physical grid in countries like Australia, we’ve always invested in market-leading analytics.

So what’s the next frontier for power price forecasting in an increasingly complex and merchant world? Traditionally, it’s been enough to model capture prices on a regional or even country-wide basis. Increasingly, however, location matters. The more renewables there are on the system, the more it becomes important to understand how your specific site is correlated with the rest of the country.

That’s why we have developed the Amun software solution to provide site-specific capture price forecasts. The tool allows project developers and investors to first perform an initial assessment of many sites using built-in data, then secondly to calculate a robust site-specific capture price using measured wind data. At both steps the tool gives results both for Aurora’s central forecast and our low/high risk scenarios.

Want to learn more how Aurora’s capture price forecasts have helped to raise financing and how we can help you? Interested in using Amun to obtain Central and P90 price projections for your specific site? Get in touch!

* Link

Author: Oliver Kerr, Senior Commercial Manager, Aurora Energy Research

Our subscription services

Our subscription service provides you with long-term market forecasts, strategic insight reports, interactive client forums and access to our market data and analytics platform.

Interested? Please sign up to receive our complimentary monthly summary reports via email. Simply tick which reports you would like to receive. Please indicate your consent in the box provided and view our privacy policy.

If you would like further information on a particular subscription service or report please enter further information in the box provided.



  • Please find our privacy policy here

  • Read our commentary, following the widespread disruption caused by power outages on Friday 9th August.

    Read More
  • With more solar and wind generation entering the energy system, BEIS and Ofgem are turning their focus onto the emergent and growing markets for flexible energy and capacities. Aurora’s commentary now published of their joint consultation: A smart, flexible energy system – A call for evidence.

    Read More
  • Aurora’s Flexibility Market Summary provide a concise monthly review of recent events in key markets for flexible technologies. The initial focus is on the balancing mechanism and firm frequency response (FFR) markets with more ancillary services markets to be added over time.

    Read More