The increasing prevalence of distributed energy is giving rise to a number of trends, disrupting the manner in which utilities operate their grids.

The biggest trend within the distributed energy space is that utilities are being forced to become much more data-driven, in terms of their approach to maintaining and operating the grid. This is demonstrating itself in several noteworthy ways and is influencing the need for ubiquitous IoT solutions, as well as communication and connectivity models.

Today, data is making its impact felt across utility operations, from energy efficiency programmes to demand response to customized consumer engagement.

Distributed, digital and disruptive

The grid is changing, becoming more nuanced and more dynamic. There is an old joke in the utility space that Alexander Graham Bell couldn’t make heads or tails of a modern-day phone but Thomas Edison could very easily operate today’s grid. Generally speaking, this was true, until now. Utilities need to be granular in how they understand loads, forecasts and changes at the customer level. It’s no longer enough to get by with big picture econometric models of the system as a whole.

Distributed energy resources (DERs) are forcing utilities to impacts down to the transformer level.

It is also important to note that DERs are acting much more like a ‘competitor’ than most utilities have seen in the past and even though ‘grid defection’ is still difficult and expensive, it is a real and achievable option. This is pushing utilities to squeeze costs out of the business wherever possible to ensure their rates are as competitive as possible. It’s also tweaking the traditional approach to asset inspection or asset replacement, which has been very cadence-based in the past, with utilities replacing or repairing the oldest assets each year fairly equally across all parts of the system.

With the introduction of AI and machine learning into utility best practices, utilities are now able to specifically target assets at risk of failure and can identify which ones can keep running for a while longer without the expense of replacement.

One of the most important trends in the distributed energy space is the generation of valuable data, which is in turn raising the question of who should own grid data. Utilities are feeling pressure from both their customers and regulators alike to prove they are being good stewards of grid data. In other words, it’s important to demonstrate a data-driven approach to asset inspection, improving reliability through more granular load forecasting, or using consumer data to better connect and engage with their customers in a one-on-one manner.

Data: The bottom line

A great example of how utilities can use a data-driven approach is for them to become more targeted in their customer engagement and participation in demand response and energy efficiency programmes. Traditionally, many utilities have utilised a personabased approach to customer engagement, with five to ten customer categories. They engage with each customer based on the profile category they fall into. All of that is changing rapidly. Data science and data availability are now allowing utilities to engage with each individual customer based on their unique attributes. In the same way that Amazon offers a customised tailored product suggestion to a specific consumer, utilities can now uniquely connect with their customers to drive significantly higher adoption of efficiency and demand response programmes, while also boosting overall customer satisfaction.

Data is also allowing utilities to run their demand response programme in a precise way, comparing what type of impact various demand response programmes will have on any given day. Furthermore, because they are now able to forecast down to the individual customer level, they can start calling demand response programmes at the circuit or substation level, instead of at the system level. This enables utilities to maximise the value they are getting from DR by providing both the normal systemlevel benefits while also starting to capture distribution benefits as well.

There are indeed challenges and changes that accompany utilities starting to count on distributed customer owned devices to provide grid services. When I was leading the “Grid of Things” team at Pacific Gas & Electric, we were working on a pilot using customer owned solar and batteries to provide distribution services through a Distributed Energy Resources Management System (DERMS). One day, we lost connection with one of the customer-sited smart inverters because the customer got rid of their WiFi, which our vendor was using to communicate with the equipment. Utilities are used to building in multiple layers of protection and redundancy; however this becomes incredibly expensive to scale to millions of DERs. As DERs become more prolific and utilities start counting on these to provide safe and reliable grid operations, adopting a new portfolio-based approach will be needed. Counting on any one specific DER will be replaced by the aggregate, ensuring a certain percentage of resources being available at any given time (and with WiFi on!) at any given time. We’ll need to use analytics where previously we hardcoded reliability into our infrastructure.


The biggest change will need to be standardisation across customer-owned IoT devices that connect to the utility. As utilities start relying on distributed energy resources to provide reliable and stable grid benefits, there will need to be improvements in terms of how these devices are connected and controlled. Imagine losing a power plant because someone unplugged the WiFi.


Tom Martin is the managing director of product management, energy & utilities at TROVE Predictive Data Science. Tom works to bring AI and data science products, known as solvers, into utilities by focusing on the actionable use cases and insights that enable changes to business practices and produce measurable value. Prior to TROVE, Tom led the emerging grid technology team at Pacific Gas & Electric: leading the implementation of new technology and analytics in support of PG&E’s electric operations as PG&E looked to reduce operational costs, improve safety, and increase reliability through a renewed focus on implementing a data-driven culture.

This article was originally published in Smart Energy International 1-2019

Read all articles via our digital magazine today