PPA Route to Market Imbalance Risk Analysis An
Description: PPA Route to Market Imbalance Risk Analysis An Update Oliver Rix 12th April 2013 Introduction Outline Methodology Historic Imbalance Prices Modelling of Probabilistic Distributions Future Imbalance Cost and Risk (On-Going Analysis)
Related Topics
Download Presentation
"PPA Route to Market Imbalance Risk Analysis An" is the property of its rightful owner. Permission is granted to download and print the materials on this website for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Presentation Transcript
slide1. PPA Route to Market Imbalance Risk Analysis An Update Oliver Rix 12th April 2013<br>
slide2. Introduction
Outline Methodology
Historic Imbalance Prices
Modelling of Probabilistic Distributions
Future Imbalance Cost and Risk (On-Going Analysis)
Questions Imbalance Risk Analysis Agenda<br>
slide3. Imbalance Risk Analysis Introduction In response to recent evidence that independent generators are finding it increasingly difficult to secure commercially viable power purchase agreements (PPAs), the Department of Energy and Climate Change (DECC) issued an open call for evidence to the industry in April 2012. This analysis supports the qualitative findings from the call for evidence responses and looks to quantify the present and future risk associated with imbalance costs and the materiality of this for generation projects. An empirical data set of historic generation volumes and cash-out prices is used to build probabilistic distributions of the key variables for different asset types to enable simulation of imbalance costs. Appropriate sensitivities can then be applied to assess potential future imbalance cost and risk. An insight into historical imbalance costs
A modelled imbalance risk for different generation types based on historic data
Future potential imbalance risks and costs for different generation types Project Introduction Analysis Goal Analysis Rational Analysis Outputs<br>
slide4. Imbalance Risk Analysis Context Market participants buying and selling physical power are responsible for their own balancing on a half-hourly basis.
To the extent that a participant’s net position based on generated volumes, customer demand and wholesale purchases and sales is not zero, this is treated as an imbalance and settled against ‘cash-out’ prices.
The cash-out price that is applied to the imbalance depends on the direction of the imbalance relative to the overall system imbalance: where the imbalance ‘helps’ at the system level, a market-related price is applied, whereas if the imbalance exacerbates the overall system position, a price is applied reflecting the System Operator cost of balancing (“System Buy Price” or “System Sell Price”).
SBP/SSPs can be at a significant and volatile premium/discount to the underlying wholesale price. GB balancing arrangements recap GRAPH – sample time series of MIP/SBP/SSP<br>
slide5. Imbalance Risk Analysis Wind asset assessment The actual imbalance accruing to participants will be a function of their portfolio and trading strategy.
We are aiming to isolate the element of imbalance that can be attributed to uncertainty in relation to the level of outturn generation from an asset.
We have used public domain data for transmission-connected assets (BM Units).
Final Physical Notifications (FPNs) represent the information on expected output provided by generators to the System Operator at gate closure, 1 hour ahead of delivery – we use these as our proxy for the forecast information.
We compare this to Metered Output, and treat the difference as a ‘forecast imbalance’.
We then calculate an imbalance cost by applying the appropriate cash-out price for that half-hour (depending on the relative direction of the forecast imbalance. Historic data set<br>
slide6. Imbalance Risk Analysis Definitions We define imbalance risk as the potential for increased costs associated with uncertainty around the expected level of imbalance cost
We propose to quantify this using a simulation model to derive a probability distribution and define a metric based on the difference between the mean (expected) and a 95th percentile worst case Imbalance Risk Imbalance Cost MV = Metered Volume (MWh)
FPN = Final Physical Notification (MWh)
MIP = Market Index Price (£/MWh) ILLUSTRATIVE<br>
slide7. Imbalance Risk Analysis Outline Methodology Define our Characteristic Generation Assets • Calculate the Probabilistic distributions for our characteristic assets’ generation volume and cost variables
• Found from empirical data sources
• For each asset type Model Historic Imbalance Price and Risk • Use our characteristic generation assets and modelled distributions to calculate historic imbalance risk and associated cost
• Check model sensitivities and calibration for different asset types and sizes Stage Outline Methodology Output Model Future Imbalance Price and Risk • Develop changes to distributions representing potential:
Evolution of balancing arrangements
Changing supply/demand fundamentals Historic System Imbalance Price and Risk • Modelled historic imbalance volumes and prices
• Identify trends in asset sizes, types, generators Wind Forecasting Insights • Investigate collated data for trends in imbalance prices and volumes
• Investigate patterns in FPNs and how to treat these in the modelled distributions Future System Imbalance Price and Risk • Modelled future imbalance volumes and prices for different scenarios and time horizons<br>
slide8. Imbalance Risk Analysis Historic Imbalance Prices Wind forecasting insights
Imbalance costs vary between generators and asset sizes
Smaller assets tend to accrue larger imbalance costs Weighted Historic Imbalance Prices and costs - 2012 By Generator By Wind Asset Type and Output Patterns of FPNs relative to metered output suggests different strategies for managing forecast uncertainty To Add: Other years Possibly remove this graph – difficult to explain! To Add: Show as % of annual revenue<br>
slide9. Imbalance Risk Analysis Probabilistic Methodology Develop Characteristic Generation Assets
For different asset types the model defines probabilistic distributions of the asset FPN, NIV and Metered Output (MO) values These are developed from a sample of empirical data with distributions calculated for ½ hourly periods, for 6 double month groups over a year Refine Model Parameters Empirical
Database Cash Out Prices Model Inputs Probabilistic Model Inputs to Future Imbalance Cost Model Modelled Historic Distributions Modelled Distributions e.g. Final Physical Notification (FPN), Metered Volume (MV), Net Imbalance Volume (NIV) Parameter Sensitivities e.g. Cash out agreements, Changing characteristic generation asset fundamentals Probabilistic Distributions For WE, WD, different generation types, optimal trading strategy Calculate Metered Volumes Calculate Implied Imbalance Price Simulate for different imbalance price scenarios Historic Imbalance Cost and Risk MODELLED HISTORIC IMBALANCE COST AND RISK<br>
slide10. Imbalance Risk Analysis Probabilistic Distributions for Different Asset Types - historic Onshore Wind Mean Imbalance Price: £2.14 /MWh Offshore Wind Mean Imbalance Price: £1.4 /MWh ILLUSTRATIVE – TO BE UPDATED
WILL STATE ASSUMPTIONS AROUND STRATEGY ETC<br>
slide11. Imbalance Risk Analysis Probabilistic Distributions for Different Asset Types - future Onshore Wind Mean Imbalance Price: £2.14 /MWh Offshore Wind Mean Imbalance Price: £1.4 /MWh REPLACE WITH SENSITIVITIES STRESSING DISTIBUTIONS OF IMBALANCE PRICES<br>
slide12. Imbalance Risk Analysis Questions?<br>
slide2. Introduction
Outline Methodology
Historic Imbalance Prices
Modelling of Probabilistic Distributions
Future Imbalance Cost and Risk (On-Going Analysis)
Questions Imbalance Risk Analysis Agenda<br>
slide3. Imbalance Risk Analysis Introduction In response to recent evidence that independent generators are finding it increasingly difficult to secure commercially viable power purchase agreements (PPAs), the Department of Energy and Climate Change (DECC) issued an open call for evidence to the industry in April 2012. This analysis supports the qualitative findings from the call for evidence responses and looks to quantify the present and future risk associated with imbalance costs and the materiality of this for generation projects. An empirical data set of historic generation volumes and cash-out prices is used to build probabilistic distributions of the key variables for different asset types to enable simulation of imbalance costs. Appropriate sensitivities can then be applied to assess potential future imbalance cost and risk. An insight into historical imbalance costs
A modelled imbalance risk for different generation types based on historic data
Future potential imbalance risks and costs for different generation types Project Introduction Analysis Goal Analysis Rational Analysis Outputs<br>
slide4. Imbalance Risk Analysis Context Market participants buying and selling physical power are responsible for their own balancing on a half-hourly basis.
To the extent that a participant’s net position based on generated volumes, customer demand and wholesale purchases and sales is not zero, this is treated as an imbalance and settled against ‘cash-out’ prices.
The cash-out price that is applied to the imbalance depends on the direction of the imbalance relative to the overall system imbalance: where the imbalance ‘helps’ at the system level, a market-related price is applied, whereas if the imbalance exacerbates the overall system position, a price is applied reflecting the System Operator cost of balancing (“System Buy Price” or “System Sell Price”).
SBP/SSPs can be at a significant and volatile premium/discount to the underlying wholesale price. GB balancing arrangements recap GRAPH – sample time series of MIP/SBP/SSP<br>
slide5. Imbalance Risk Analysis Wind asset assessment The actual imbalance accruing to participants will be a function of their portfolio and trading strategy.
We are aiming to isolate the element of imbalance that can be attributed to uncertainty in relation to the level of outturn generation from an asset.
We have used public domain data for transmission-connected assets (BM Units).
Final Physical Notifications (FPNs) represent the information on expected output provided by generators to the System Operator at gate closure, 1 hour ahead of delivery – we use these as our proxy for the forecast information.
We compare this to Metered Output, and treat the difference as a ‘forecast imbalance’.
We then calculate an imbalance cost by applying the appropriate cash-out price for that half-hour (depending on the relative direction of the forecast imbalance. Historic data set<br>
slide6. Imbalance Risk Analysis Definitions We define imbalance risk as the potential for increased costs associated with uncertainty around the expected level of imbalance cost
We propose to quantify this using a simulation model to derive a probability distribution and define a metric based on the difference between the mean (expected) and a 95th percentile worst case Imbalance Risk Imbalance Cost MV = Metered Volume (MWh)
FPN = Final Physical Notification (MWh)
MIP = Market Index Price (£/MWh) ILLUSTRATIVE<br>
slide7. Imbalance Risk Analysis Outline Methodology Define our Characteristic Generation Assets • Calculate the Probabilistic distributions for our characteristic assets’ generation volume and cost variables
• Found from empirical data sources
• For each asset type Model Historic Imbalance Price and Risk • Use our characteristic generation assets and modelled distributions to calculate historic imbalance risk and associated cost
• Check model sensitivities and calibration for different asset types and sizes Stage Outline Methodology Output Model Future Imbalance Price and Risk • Develop changes to distributions representing potential:
Evolution of balancing arrangements
Changing supply/demand fundamentals Historic System Imbalance Price and Risk • Modelled historic imbalance volumes and prices
• Identify trends in asset sizes, types, generators Wind Forecasting Insights • Investigate collated data for trends in imbalance prices and volumes
• Investigate patterns in FPNs and how to treat these in the modelled distributions Future System Imbalance Price and Risk • Modelled future imbalance volumes and prices for different scenarios and time horizons<br>
slide8. Imbalance Risk Analysis Historic Imbalance Prices Wind forecasting insights
Imbalance costs vary between generators and asset sizes
Smaller assets tend to accrue larger imbalance costs Weighted Historic Imbalance Prices and costs - 2012 By Generator By Wind Asset Type and Output Patterns of FPNs relative to metered output suggests different strategies for managing forecast uncertainty To Add: Other years Possibly remove this graph – difficult to explain! To Add: Show as % of annual revenue<br>
slide9. Imbalance Risk Analysis Probabilistic Methodology Develop Characteristic Generation Assets
For different asset types the model defines probabilistic distributions of the asset FPN, NIV and Metered Output (MO) values These are developed from a sample of empirical data with distributions calculated for ½ hourly periods, for 6 double month groups over a year Refine Model Parameters Empirical
Database Cash Out Prices Model Inputs Probabilistic Model Inputs to Future Imbalance Cost Model Modelled Historic Distributions Modelled Distributions e.g. Final Physical Notification (FPN), Metered Volume (MV), Net Imbalance Volume (NIV) Parameter Sensitivities e.g. Cash out agreements, Changing characteristic generation asset fundamentals Probabilistic Distributions For WE, WD, different generation types, optimal trading strategy Calculate Metered Volumes Calculate Implied Imbalance Price Simulate for different imbalance price scenarios Historic Imbalance Cost and Risk MODELLED HISTORIC IMBALANCE COST AND RISK<br>
slide10. Imbalance Risk Analysis Probabilistic Distributions for Different Asset Types - historic Onshore Wind Mean Imbalance Price: £2.14 /MWh Offshore Wind Mean Imbalance Price: £1.4 /MWh ILLUSTRATIVE – TO BE UPDATED
WILL STATE ASSUMPTIONS AROUND STRATEGY ETC<br>
slide11. Imbalance Risk Analysis Probabilistic Distributions for Different Asset Types - future Onshore Wind Mean Imbalance Price: £2.14 /MWh Offshore Wind Mean Imbalance Price: £1.4 /MWh REPLACE WITH SENSITIVITIES STRESSING DISTIBUTIONS OF IMBALANCE PRICES<br>
slide12. Imbalance Risk Analysis Questions?<br>