Analysing the data flood: Opportunities and challenges in a changing energy market

by Matt Nidd, Utilities Principal, DB Results

As big data plays an increasingly important role in how utilities operate and plan for the future, Matt Nidd considers some of the unique technologies in the market that can help utilities get the most out of their data.

Industry context

A number of relatively recent or impending changes in the electricity sector are tightening the ability of distribution network owners to recover regulated revenue through traditional methods such as asset enhancement and replacement.

These changes include:

• The inception of the National Electricity Customer Framework, driving a requirement to be customer rather than premise centric. Customers must now, more than ever, remain informed of potential disruptions in energy supply in a manner that recognises them and their previous interactions with the organisation.

• Recent reductions in energy consumption (excluding the last Victorian winter!), creating a perceived over-
investment in the network.

• The pending introduction of the range of Power of Choice initiatives from the AEMC, removing the guaranteed ownership (and therefore return) on mass market meters, and necessitating the decision on whether to compete with agile new entrants to the metering market.

The opportunity landscape 

These disruptive changes obviously assume no compromises around quality of supply or safety, making maximising the useful life of an asset more important than ever. Simply sustaining a reliable network is no longer good enough.

Further up the value chain, Power of Choice creates a new call­to­action and ongoing customer conversation for retailers that can only be fully taken advantage of if the intersection of customer value, location and appropriate bundled offers are fully understood.

Distribution businesses, information and communications technology professionals, and operational technology managers must consider how to respond to the above changes in the context of the data available from their existing transactional systems.

The proliferation of smart meters in Victoria has provided a unique opportunity to extend the understanding of both retail and distribution participants when it comes to customer behaviour and asset utilisation.

The new data that will become available with the retailer­led Power of Choice rollout is by no means the only source of insight into customer behaviour. There remains many other sources of data that, when effectively overlaid, can provide valuable insights.

Business insights from big data

In assessing opportunities to provide actionable business insights, decision­makers must balance the time it will take to implement a long run information management strategy (IMS) against often uncertain business requirements, and the time to value ratio that this approach involves.

Establishment of an information strategy, procurement of the necessary data storage and analytics capabilities, and integration of the solution and associated governance framework is likely to take upwards of two years and many millions of dollars before the first use cases are enabled. Often the business will have grown impatient over this period and sought the data via alternative means (working outside of the information strategy and probably information and communications technology!), or the requirement will have changed, materially rendering the first use case delivery redundant.

Due to the relatively early stage in the technology lifecycles of the capabilities mentioned here, two years is a very long time, and software and hardware may have been partially or completely superseded during the life of the program. All of these factors must be considered carefully to ensure a sustainable business case and realisation of business outcomes.

A mitigation for the potential pitfalls of a long­running IMS implementation is to invest in a ‘two speed’ approach to the delivery of business insights. This involves considering complementary technologies, such as the emerging group of solutions which provide ‘situational insights’ as part of an IMS.

These tools provide relatively rapid implementations through existing, industry­specific use case and out of the box adapters (with a timeline of 6­12 months to value, instead of two years) and fast off­premise deployments. The insights these solutions provide are often based on a combination of subscriptions to existing transactional data services (via an existing service bus) and other operational data sources.

Due to this approach, the insights lend themselves to operational as well as more strategic uses, as they are situational or current, rather than based on very long­term data analysis. With a strong set of guiding principles, the utilisation of these sorts of tools can be complementary (rather than tactical) to the solutions delivered through the longer running implementation approach running in parallel. In fact, some businesses are finding that a second analytics tool is redundant as they learn to fully leverage the capabilities of the situational toolsets and existing use cases.

In developing an IMS, consideration must be given to the range of tools available and the likely lead time the selection of strategic tools will have on the realisation of business outcomes. The selection of tools and the implementation overhead must be weighed up against the regulatory period the value needs to be derived for/from; as well as the business’ ability to endure a long running implementation before initial business outcomes can be realised. Fortunately there are now more options than ever to ensure businesses can leverage their data progressively with innovative approaches to asset and customer management, in line with regulator and customer expectations.

In conclusion, it’s worth reflecting on some exciting use cases that are currently commercially available, utilising the approaches outlined above.

Distribution

• Predictive analysis utilising power quality data from smart meters or other sensors for:

» Early fault detection

» Theft detection

» Impacts of solar density on network assets.

• Predictive analysis (presented spatially and as potential events) of fire danger to network assets that utilises existing asset information, weather data and historical trends. This is being used to position fault crews strategically on high fire danger days and to inform customers of potential risks to their supply.

Retail

• Analysis of customer value, geographic location (and density) and churn statistics to establish targets for retail Power of Choice bundles, and the appropriate software and meter communications technologies for servicing those customers’ requirements.