The Mechanics of Grid Edge Intelligence

Emerging trends – electric vehicles (EVs), distributed energy resources (DER), advanced battery applications, smart devices and vehicle-to-grid (V2G) technology – have fundamentally transformed how we design, operate and maintain the grid.

A decade ago, utilities were tasked with adding intelligence to their transmission and distribution substations; today, utilities are adding digital intelligence to the so-called “grid edge” to connect with the smart devices located there. Some are becoming prolific on the customer side of the meter (e.g., EVs, smart thermostats, appliances). These new DER are accelerating us past the days of a single-direction, centralized electric grid and towards a bi-directional, distributed grid.

This acceleration amplifies the need for advanced communications networks as the industry moves towards a smarter, more advanced grid.

Adding Intelligence to the Grid

Grid edge refers to the proximity of the grid to utility customers at their homes and businesses – whether they are 150’ away from a substation, or seven miles away. The grid edge is not associated with traditional generation or transmission. Accurately defining the grid edge depends on the system topography, the load profile and customer demographics – and it will change over time.

COVID-19 has already driven a significant portion of residential customers to work from home, changing load peaks across the U.S. A recent IEEE report states: “The biggest limitation of such an analysis comes from the lack of available higher-resolution data on electricity consumption. Each of the major regional transmission organizations publishes power load and price numbers daily for their electricity markets, but this reflects a fairly large geographic area that often covers multiple states.” Armed with grid edge intelligence, utilities will be able to better understand these trends down the road – and more importantly, understand the value of this data.

When it comes to adding intelligence to the grid edge, utilities can pursue a centralized intelligence approach, a distributed intelligence approach, or a hybrid approach. Generally, centralized intelligence at the headend is considered more “intelligent” to grid conditions because of a more holistic, global view, while distributed intelligence is more sensitive to fast change because everything occurs at the grid edge.

Centralized intelligence involves simpler devices talking to the headend using protocols like DNP3 (IEEE 1815). Headend analytics require reliable communications to the edge, although issues can arise around late or missing data. More polled data will be transmitted to the headend for analytics and response, so polling is less on-demand. The benefit is that the headend analytics typically have better insight into the system and adaptable logic.

Distributed intelligence involves complex devices at the grid edge with analytics and peer-to-peer communications. The disruption of grid edge analytics relies on peer-to-peer communication, typically low latency, with less polled data traveling back to the headend, and more data available on-demand. Edge analytics typically have reduced system perspective and more hard-coded logic. The benefit is the lightning-quick reaction of the grid edge devices even when communications to the headend may be cut or compromised.

A hybrid option offers a combination approach, with both advanced distribution management solutions (ADMS) and grid edge intelligence. This approach balances the need for more oversight (distributed) with the need for high initial investment (centralized), providing quality data to drive decision making.

Understanding the Impacts

As the grid edge becomes populated with thousands and thousands of devices, utilities must understand the high-level impacts of grid edge intelligence.

Cybersecurity is a critical component of any grid edge strategy. Looking at the distribution automation industry, many OEMs have not adopted the common cybersecurity measures seen on an IT system today – e.g., user controls, passwords, password controls, logging controls, encryption. While this is improving, utilities must demand that OEMs implement cybersecurity controls.

Click image to enlarge

Figure 2. High-level control concepts are necessary to manage devices at the grid edge
 

A converged field area network (FAN) allows one team for management and support, versus multiple disparate groups and technologies. Education is key to enabling the team to perform maintenance and troubleshooting on one converged network that supports the needs of many different groups.

Beyond education, utilities need to adapt processes and procedures. For example, utilities that choose a centralized architecture may need to perform significant work to implement an ADMS and the procedures that work with the ADMS and grid edge devices. This change will cascade down to the utility’s line crews, operations team, control centers and engineering teams.

This may result in traditional skills falling to the wayside. Utilities once had very defined teams – e.g., the distribution engineers, the planning engineers, etc. The utility may need to work with change management groups to understand the necessary skillsets going forward, and how those needs will impact any union operating agreements.

A converged network must be robust enough to handle the applications needed today, while also providing a firm foundation for the future. Once built, will the network support future additions like gas devices, transmission devices, or energy supply devices?

Lastly, utilities must implement a robust asset management program to help them manage thousands of distributed intelligence devices, all of which must be managed to cybersecurity standards, including access controls and configuration changes.

6 Steps to a Smarter Grid

So how do utilities enable grid edge intelligence? Adding applications requires a deep understanding of where intelligence is added. The mechanics of grid edge intelligence provides a proven, well-defined roadmap for a project that addresses the following concepts:

1.      Planning. Identify and gather together all utility stakeholders. Establishing a common understanding and definition of the project will be key to determining roadmap success.

2.      Assessment. Identify and document the current state throughuse cases. What are the current state applications? How are the data and security designed? Is there enough staff? Do they have the necessary skills? Is there dependency on other providers such as equipment vendors? How much infrastructure is owned versus leased? These are all critical questions.

3.      Architecture. Create a high-level conceptual design for the future state, considering which solutions are commercially available – and affordable. Design an architecture that’s going to last 15 to 20 years. Consider equipment life cycles but don’t focus on them; instead, design the architecture to serve as a framework and adapt as new technologies emerge. Document the performance and other requirements. Discover impacts on utility processes. Determine project metrics and required testing before the design phase. Reach out to vendors to start the procurement process.

4.      Design. Use theavailableinformation and testing results to create detailed functional requirements. Select the vendors and create design standards so there is a complete bill of materials that meets all requirements. Create detailed designs and start training and process improvements in tandem with procurement. Finalize the numbers – e.g., capital costs, operating costs, etc.  

5.      Implementation. With stakeholder buy-in obtained, all system requirements tested, new processes designed, and completed designs and metrics in place, a well-organized, detailed plan for implementation is needed. Consider which impacts provide the best overall value while minimizing risk and plan accordingly. Coordinate staffing resources and training.

6.      Optimization. Once implementation is in process, optimization may begin. Refine, improve and fine-tune adjustments to the selected applications, devices, processes and converged FAN based upon the selected metrics.

If done correctly, this roadmap process keeps the journey to grid edge intelligence moving. It is flexible: utilities can start work on the conceptual design even as they start the planning process; they can line up for implementation and check boxes even as they work through the architecture phase. Identifying gaps early will help ensure that the detailed design, implementation and optimization phases run smoothly.

When it comes to enabling grid edge intelligence, no matter what approach a utility wishes to take – centralized, decentralized or a hybrid – they must gather their application requirements, understand their current state, and plan a roadmap. Placing the grid edge at the forefront of the roadmap allows utilities to gain valuable insight, enabling them to stay innovative and agile, safely accelerating them towards a more sustainable, resilient and reliable future grid.

Black & Veatch