Endeavour Energy manage 3million assets servicing 2.7million people in NSW. Over the past 18-24 months we have overhauled our traditional annual qualitative view of asset intervention into a modern quantitative data driven multi-year risk-based approach.
Individual assets are now assessed and identified for intervention to maximise customer value, create alignment with strategic objectives and provide risk outcomes for differing proposed investment scenarios.
The new framework provides a transparent consistent evaluation approach across our internal engineering, regulatory, finance and delivery teams. Optimised asset management scenarios are created based on each stakeholders’ constraints, further allowing improved communication with external stakeholders including customers.
Endeavour Energy owns a vast range of assets with a wide variety of functions, performing under a range of environments. Individual assets can have significantly different service levels, lives, maintenance and inspection programs, all impacting the probability of an aged-based wear-out failure occurring.
Asset risk, due to functional failure varies across our network by individual asset. This variation may be due to factors such-as the asset’s location in the network, impact to other local assets, amount of energy disrupted and time to restore, negative environmental impacts as well as public/worker safety.
Our objective was to develop and implement a repeatable modern quantitative data driven multi-year risk-based asset management system displaying best practice that evaluated each of our 3 million assets. Intervention decisions where to maximise customer value taking into consideration real world contrast and limitations.
Figure 1 – Summary of tasks required to provide asset type data to feed into the new risk-based economic assessment
Through the achievements of our efforts, the asset risk methodology is now consistently applied across all assets to ensure a balanced assessment. The cost-benefits of all assets and their possible investment options are calculated and prioritised in order of its optimal intervention timing and total risk-cost benefit.
Investment scenarios show maximum risk-cost benefit at time of optimal intervention or for any given investment timing, adjusted or constrained across the full portfolio of assets. Asset decision investment outcomes are now supported through traceable line of sight right back to our companies’ overall strategic objectives.
The application of the system has been rolled out and adopted by a range of users from different parts of the business including engineering, regulatory, finance and delivery teams. The process has been developed to be transparent, easily repeatable, and flexible to account for a real-time operating environment to allow for new emerging risks, revised objectives, expenditure or work force adjustments or other real-world constraints.
This required a significant systems and cultural overhaul across our business coming away from our traditional 12-month qualitative view of identifying specific assets for intervention and delivery approval to a modern risk-based quantitative data driven view identifying 5-10years of future specific assets where each had been economically justified and supported for its optimal intervention.
- Understanding our data
Initial task was to understand and use the data we had available. Data within our organisation is located across many incompatible legacy systems as well as newer SAP system undergoing various phases of implementation. Data within these systems consists of millions of historical and in-service snippets of data. First challenge was to bring all this data together, sift, cleanse and tease out something meaningful.
To assist in evaluating this data we used standard off the shelf data integration tool called FME and built data flows to combine information from these multiple internal data sources as well as external sources to create the most accurate live view of the network possible.
Figure 2 – In-house workshop training ourselves to use the FME data integration tool
- Asset hierarchy
An asset hierarchy was created where the asset base was initially grouped into similar asset classes. Particularly with high volume low value assets. Asset types with like characteristics were also further merged under their respective asset classes.
- Asset Level Probability of Failure Curves
All available historical asset failure data (conditional and functional) was extracted and processed at an asset class level to develop probability of failure curves for assets that allowed each asset to develop a relationship between age / condition and its likelihood of failure occurring both today and every year into the future. This approach predominantly used Weibull and CNAIM reliability modelling methods.
The curves were tested back against the asset population to ensure real world outcomes matched those of the model.
Figure 3 – Asset data used to calibrate Weibull parameters
- Asset Level Likelihood of Consequence values
Financial data associated with each asset failures was combined with asset failure date and geospatial data related to the asset to gain an understanding of the likelihood of different consequences occurring.
Figure 4 – Example of CoC geospatial layers developed to assess asset location compared to scale of risk
- Cost of Consequence Models
Users can apply any of the applicable six customer value consequence models developed. These included:
- Legal and Compliance
Geospatial analysis of each individual asset using both internal and external data sources were combined in conjunction with network models to understand the true impact to customers of any single asset failing on the network for each of these measures.
Figure 5 – Example of the typical visual output of our models developed in-house
- Economic Evaluation tool – Asset Risk / Cost Model
Development of an economic model to calculate both net present value and cost / benefit analysis of a proposed investment or investments against a base case “run to failure” scenario. Considering the ultimate survivability of the original asset (if an investment was selected) and the reduction in “true” benefit to customers.
Maximising customer value understanding our customers’ needs and risk costs and prioritising our investment portfolio to best maximise their needs.
This risk-based evaluation identified an asset replacement cost of over $1billion worth of available condition-based scope for consideration with the FY25-FY29 period. Through this risk-based process this was optimised to $575 million, at a cost saving to the customer as Endeavour Energy manages the underlying risk within its current lifecycle programs.
- Asset Management Outcomes
Documenting our findings resulted in the development of:
- x3 System strategies: to explain where we are going to focus;
- x13 Asset class plans: that outline current asset performance delivery; and
- x54 Cases for investments: identifying specific investments that are robust and defensible.
Figure 6 – Identifying risk outcomes based our strategic positions
- Asset Risk Methodology
Ensuring consistency was key to successful outcomes. This included a fully documented process for the application of the asset risk methodology and support a new common language between the different parts of our business from board level through, engineering out to our delivery teams in the field.
Figure 7 – Asset risk cost modelling framework
- Customer reviewed
One of our final challenges was communicating to our customers in a language (risk-cost) which mums and dads can understand. We put all our proposals in front of our customers across multiple meetings. The audiences for these meetings included:
- 100 residential and small business customers;
- Informed customers such as developers, local councils and interest groups; and
- Regulatory reference group and informed customer advocates.
Figure 8 – Example of one of our interactive customer advocate meetings
Recent feedback as a result from these in-depth customer engagements and from our regulator shows an endorsement to our new revised approach towards asset risk-cost management and our expenditure forecasting.
In hindsight we may not have grasped the enormity, difficulty, or importance of what we set out to achieve. The capability to perform quantitative risk-based asset management analysis on 3million assets, in a consistent and transparent manner opened many more opportunities than first contemplated.
Whilst the broad objective may have been to provide risk-based quantitative justification in accordance with the regulator’s guidelines and framework, what was achieved was much more.
The ability to openly communicate the change in network risk (whether it be safety, reliability or environmental) and future trends in each of these risk measures based on proposed decisions is a powerful tool and arguably the most important achievement from this body of work.
The conversations that stem from the data created by this framework, initiated true asset management discussions across different parts of the business. Examples of the insight obtained from this framework are:
- An in-depth understanding of the difference between risk, likelihood and consequence across the network in a spatial context (e.g. does one area have a greater risk than another). This insight allowed the risk profile across the network to be normalised.
- Comparison of the individual risk measures (reliability, safety, environmental etc) across either an asset, a region or the network. To understand what is needed from an investment point of view to move any one of these risk measures and what are the true drivers for investment.
- Discussions if current risk levels are acceptable, should we be aiming to increase or decrease them.
- If a new constraint appears (labour, financial or procurement) how best is the portfolio remixed to ensure organisational / customer objectives continue to be best met. What impact will the constraint have on future network performance and risk (e.g. will it result in a future increase in unassisted asset failures, cost and risk).
To bring this framework together the organisation needed to combine several skillsets including asset management, financial acumen, adaptable risk appetite and data literacy. No single individual or team can have the skills needed in each of these areas to successfully complete this project. Hence a great culture and strong communication channels across the business were needed that felt comfortable to take risk, challenge each other and ask for help were needed in order to achieve what was accomplished.
Since starting to communicate the process developed, the outcomes that are achievable and illustrate the flexibility and transparency of the framework, comments including “is this level of understanding real?” and “Endeavour Energy are the first organisation in the industry to truly have a larger justified portfolio than what it is being proposed to be done” have been received.
Whilst we may have set out to perform a detailed quantitative risk assessment on each asset, the power of the data to transform asset management from a technical language only understood by asset management professionals to something that can be translated for a variety of stakeholders across the organisation and externally, from Boards to Customers is the real achievement.