Converging IT and OT data for intelligent operations

Why real-time access to critical downstream operational data and utility commercial IT systems is key to making informed business decisions.
Published: Wed 26 Sep 2018

Grid monitoring is an exciting development for utilities, granting them opportunities for operational efficiency, new services and better preparedness for disruptive technologies such as distributed energy resources (DERs) and electric vehicles (EVs)

However, to get to the point where any and all of these opportunities are possible, let alone cost-efficient, data is the key.

Transforming data into an asset value utilities requires analytics and near real-time processing, and is only slowed down by colocation strategies such as distributed file systems, data lakes and centralised data warehouses, where data needs to be duplicated.

One solution is to leverage a data virtualisation layer which pulls disparate data into a transaction platform, and such was the aim of the partnership between analytics and data platform providers, PowerRunner, and OSIsoft’s PI System, an operational data management system.

In an Engerati interview with Dan Garvey, Sales and Marketing Director at PowerRunner and Kevin Walsh, Global Industry Principal for T&D/Smart Grid  at OSIsoft, we learned more about the partnership supporting utilities’ operational decision management with near real-time access to critical downstream operational data, and the technology underpinning it.

Distribution network operator to system operator (DNO to DSO)

With decades of experience in the industry, PowerRunner’s principals began its journey in 2007 as a consulting company providing business and systems consulting to the independent system operators (ISOs) across the US. In the process, they identified a movement in the industry where the same business structure that was occurring at the wholesale level was going to drop down to the distribution level in the form of DERs and other disruptive technologies. Akin to the ISO need for real-time transactional data, distribution system operators (DSOs) need to know what’s happening on their grids, and they need to know in real time.

Walsh says: “If you look at what’s happening in the UK, many distribution network operators (DNOs) are transforming into distribution system operators (DSOs) over the next five years. Unless they bring in some skilled people that understand wholesale market transactions, financial settlements that take place on the transmissions side plus the volume of data that will be transacted on the distribution network they may not be prepared.”

With this change, the new volumes of data and requirements such as price with EVs and DERs will create significant difficulties, says Walsh. “The amount of data that they’re going to have to manage and digest and drive value from needs a scalable, configurable environment to futureproof them in the face of massive change.”

So, in 2012, PowerRunner pivoted into a software company and began rolling out a forecasting solution, and now a full transactional data platform. The PowerRunner Energy Platform is a data virtualization layer designed to integrate real-time operational data and commercial data with full data governance and security and highly configurable presentation layer.   

By building an adaptor which can integrate directly into OSIsoft’s PI System data infrastructure, the PowerRunner Energy Platform can leverage data virtualisation technology, and combine the service point level IT data with the OT data that the PI System collects and acts upon on the operational side of the house.

One use case which Garvey finds particularly applicable to this DNO-DSO shift is that of micro-forecasting at the service point level, citing a case study with a local utility in St Louis MO, with 2.6m metering points. He says: “We have been able to create a unique forecast model for each meter using our machine learning algorithms. As weather changes and the load changes in response to weather, for example, that tweaks our models and continually tunes them to improve over time. The temporal and spatial granularity of our forecasts provide the DSO with near-real-time situational awareness.”

“This has a lot of applications in the future DSO model. The network is going to become so dynamic and much more localised that having visibility with short-term forecasting on a circuit by circuit basis is going to create that situational awareness.” This awareness is available in the form of specific solutions for grid operations, systems planning and revenue analytics. Garvey says the solutions add value to the enterprise’s data, making analytical power available to an array of users across the enterprise.

Making data efficient to empower business decisions

With mounting pressure and expectation for utilities to make decisions more effectively and quickly, the partnership between OSIsoft and PowerRunner strives to streamline data collection, management and visualisation.

Walsh explains: “A lot of time is wasted on organising and preparing data to make it available - we remove that aspect so you can take action on it locally.”

Garvey provides further insight: “There’s a lot of organisations right now that are trying to employ a centralised data management system, but we feel the value of data is eliminating silos by putting data in the hands of the business user whether in engineering, accounting, IT or the executive suite - all of whom need an accurate window on their grid.”

He continues, explaining that creating a data structure and presenting it to a business for their consumption is easier than a data scientist searching through data to understand what a business wants: “You can’t take 20 years of business acumen and put it in the head of a data scientist so that they know what to look for.”

Instead of replicating data from system to system, data is organised and available locally in a user-friendly way for business units to utilise. Walsh says: “Electricity is real-time, therefore the utility business is real-time. The more latency you put into data processing, you’re losing the effect of responding to the data (i.e. events) in real-time.”

Garvey picks up on this thread: “Every piece of data that we pull into our environment is aligned with other disparate data sources and time stamped so that we understand the relationship between all of these various data sets and their attributes based on their historical situation.”

PowerRunner has able to take the network topology and ensure all of the systems it’s supporting are in alignment when it comes to activities such as reporting, calculating, aggregations or market settlements. Maintaining accuracy not only on the topology, but also from a financial perspective is critical in situations such as settlements.

Garvey says: “If a month from now, you want to go back and understand why there was a voltage sag or a transformer overloaded, you need to recreate the historical relationship of not only the assets, but all of the measurements that occurred at each asset on an hour-by-hour basis. Because we have the capability of time-stamping and versioning all of the data and maintaining those snapshots, you have a historical sort of ‘time machine’ which is important from both an operational and financial standpoint.” Garvey said that the solution is proven and users can save large sums of money through improved use of assets, while improving service and reliability.

IT-OT convergence and data

A pivotal issue which the OSIsoft-PowerRunner partnership addresses is the IT-OT convergence for utilities. The ability to gather OT data and time series sets, and join it with the ongoing commercial data (market pricing, rate structures, etc.) provides a holistic view of the emerging DSO model.

This also has significant benefits for cybersecurity concerns, as Walsh explains: “Because of cybersecurity needs, we have to insulate operations from corporate, so PI has become a nice security barrier between operations and getting that valuable and important data out to those who can take advantage of it.”

Key to moving forwards throughout all of this change, Walsh says, is future-proofing: “Human nature always looks at the end results, but most people forget that you have to look back to understand it. Just having that data organised in a fashion where it’s easily accessible to a business unit and the people in that unit pay dividends.

What most of our customers do with PI is make it the defacto OT data store for real-time monitoring and support.  As the utilities add more systems, devices and technology, that all have their own unique head-end system (HES), PI is the one common technology that makes the OT data available to all departments, people and systems, such as the PowerRunner Transactional Data Platform and others.