1. Project Summary
Horizon Power’s energy network currently powers over 48,000 customers in remote and regional WA. Potential weather event changes due to climate change pose an unknown risk to Horizon’s assets and ultimately to customer electricity supply. Understanding this risk will influence long-term operational and strategic asset management decisions.
In collaboration with KPMG, Horizon Power developed a ground-breaking stochastic model to quantify this risk. The model combines catastrophe and chronic weather modelling supplied by Risk Frontiers with energy distribution network dependency and asset restoration modelling. It forecasts financial and non-financial impacts of acute (bushfires, floods, cyclones) and chronic (heat, wind, humidity) risks.
2. Project Description
2.1 Asset Management Principles
Output Focus and Capabilities
When electrical assets, such as a power pole, fail due to a climatic event, there is a financial cost to repair or replace the asset. More importantly, there is a period where Horizon Power’s customers are left without power. Whilst existing standards provide adequate risk mitigation in historical and current climates to inform asset planning and budget cycles, the potential impacts of climate change on above ground assets in Western Australia required further investigation.
Our project maintained a focus on quantified outputs by assessing physical climate risk through a mix of financial and non-financial indicators. For every simulation, the metrics below were forecasted:
The non-financial metrics ensure that a high-level focus was placed on outcomes for Horizon Power’s customers. These metrics align with Horizon Power’s ambition of delivering energy solutions for regional growth and vibrant communities, through the provision of robust and resilient electrical assets and supply to mitigate the projected impacts of climate change.
Horizon Power worked closely with KPMG to collate all data and assumptions required to calculate these metrics.
The stochastic model captured the ‘volatility’ of acute and chronic weather events and the impact they cause. Quantifying the volatility associated with the risks allows Horizon Power to understand, assess, and potentially mitigate that risk. The results provide flexibility to analyse impacts across many forecast scenarios.
Climate Scenarios (based on greenhouse gas intensity):
- Representative Concentration Pathway (“RCP”) 2.6 – global warming is kept below 2°C, in-line with the ambition of the Paris Climate Agreement.
- RCP 8.5 – a fossil fuel intensive future, warming may be >5°C.
Return Periods (“RP”)
- The return period corresponds to a percentile of the distribution of the metric. For example, there is a (1/100) 1% likelihood that the outcome would exceed the 1-in-100-year return period result.
- The model stochastically produced 10,000 simulations per risk type. This provided flexibility to interpret any return period up to a theoretical 1-in-10,000-year event. This flexibility allowed an understanding of both the expected impact of the risk and the tail risks presented by climate events.
- The model is flexible to forecast results for different future years. Currently, results have been analysed for projection years 2020, 2030, and 2050.
The impact of physical climate risks on assets are changing and will continue to change. Prior to the delivery of this project, the extent of these risks had not been quantified. This project provided tangible insight into that unknown.
To action these learnings, the final report from this project was disseminated to key stakeholders in the business, including senior management. Its insights will be used to influence informed asset management decisions including:
- Asset replacement projects (eg replacing old infrastructure in high risk areas).
- Resilience building (eg replacing wooden poles with steel and concrete poles in areas of increasing bushfire risk).
- Restoration/repair workforce management (eg relocating repair resources to highest risk areas and work force escalation processes).
2.2 Originality and Ingenuity
A typical physical asset risk assessment would be conducted by insurance providers, and only if those assets were insured. Insurers have used catastrophe models in their pricing processes for decades. The limitations of this approach is:
- The analysis focusses on short term horizon based on the current frequency and severity of weather events, rather than long term forecasts.
- It focusses exclusively on asset repair and replacement costs and does not consider the impact of lost asset capability on consumers.
The challenge of combining both the financial and customer impacts of climate risks demanded an innovative and advanced modelling solution. To our knowledge, this is not something that has been previously done to the precision and robustness delivered. The key innovative elements of this model include:
- Network dependency calculations: If a distribution asset is damaged by a weather event, that asset is not affected in isolation, it also can cause downstream energy assets to de-energise. The model creates a dependency ‘map’ by identifying the key ‘feeder’ assets for every asset in the network and if a feeder is damaged, all its dependents are assumed to be de-energised.
The flow chart below depicts the logic flow of the stochastic model:
- Integrating multi-disciplinary models: The solution combined elements of traditional catastrophe models, global climate models, dynamic financial analysis models and the network dependency model described above within a stochastic framework to leverage the collective powers of each of these approaches.
The schematic below depicts the data and information flow modelled within the stochastic model:
The ingenuity is also driven by the high-quality catastrophe modelling by an experienced team of climate scientists within Risk Frontiers.
2.5 Project Management
This project was driven by Horizon Power and involved input from a wide variety of stakeholders and SMEs from within Horizon Power, KPMG Actuarial, KPMG Engineering and Risk Frontiers.
Managing expectations of stakeholders from different fields of expertise and maintaining buy-in from all parties can be a significant challenge. The key project management features which ensured the project’s success were:
- A dedicated project manager to coordinate touch points between stakeholders.
- Clear, early definitions of model processes and objectives.
- Regular feedback from stakeholders on work-in-progress reports, with a focus on ‘reality checking’ the modelling scenarios and outcomes.
The iterative feedback cycle ensured all stakeholders remained engaged and their collective insights were represented in the project deliverables.
2.6 Organisational and Community Impact
As outlined previously, this analysis will not only shape understanding but also directly impact key policies. Any organisational impacts (reduced asset replacement costs, reduced outages) will have implications for the community through reduced electricity costs and a more reliable power supply, offset by investment costs used to enhance asset resilience.
There is also a positive community impact generated from climate change impact analysis that is shared through climate related disclosure reporting. The more detailed understanding individuals and business have about the threat posed by climate change, the more support for environmental action increases. This area of research is an important mechanism to drive change.
Horizon Power provides electricity to remote and regional Western Australia spanning the largest service area of any Australian energy utility as well as the most geographically and ecologically diverse regions. Future climate projections indicate regional Western Australia, already exposed to a harsh and challenging climate, will face significant and complex changes across its vast area. The Asset Vulnerability Assessment project provides quantification and digitisation of geospatially specific risks spanning temporal and climate projection horizons for our network and generation assets in the Exmouth area. This was a complex undertaking requiring a mutli-faceted approach, that varied significantly from a standard approach to catastrophe modelling as the focus was strongly weighted to social parameters, primarily protection of customer supply. The outcomes of the assessment are being used to inform future network and generation planning, capital budget expenditure and workforce planning to ensure that we support asset and community resilience into a uncertain future climate.
Manager Sustainability, Horizon Power
4. General Comments
4.1 Project Deliverables
As part of the delivery of this project, KPMG provided the following deliverables to Horizon Power:
- A ‘Climate Change Impact Analysis’ Report: This report outlines the key findings of the analysis in a clear summary to be distributed to stakeholders including senior Horizon Power management.
- A ‘Physical Vulnerability Model Playbook’: This document provides a detailed step-by-step guide to the technical methodology used in this project. It includes explanations of what data is used, how stochastic peril events are selected, and how the calculations of each metric are performed for each simulated event.
- An analysis dashboard: This presents results across all the key outcome metrics, with the ability to filter by Acute or Chronic Weather, Future Climate Scenario, Forecast Year, Return Period, and Modelled Metric.
In combination, these three deliverables provide:
- Immediate cut-through to senior stakeholders who drive asset management policies.
- Understanding of the technical model approach to empower confident decision making.
- The ability to deep-dive into the findings and perform analysis on specific perils or asset types.
Together, these deliverables will maximise the impact of the project on Horizon Power’s asset management policies.
5.2 Future Developments
Climate change is a new frontier of risk where understanding is improving over time. The approach that we have developed in collaboration with KPMG, supported by climate data delivered by Risk Frontiers, is a continually improving and business-integrated model.
There is significant potential for development, which may include:
- Including the assets and interdependencies of more Distributed Network Supply Providers (“DNSPs”).
- Including more real-world mechanisms such as the prioritisation of asset restorations and the travel time of service teams to reach the failed assets.
- Including supply chain modelling, which would come into consideration when significant asset replacement occurs.