The advent of electric grid modernization has made clear the foundational role that information and communications technology or ICT will play in our power system as we move ahead. The upside is that the grid is becoming more thoroughly monitored and controlled from “end to end.” One big unknown is that we need to figure out how we will manage and exploit the unprecedented quantity of data – the proverbial “Big Data” – that will result.
We use quotation marks on “end to end” because although monitoring and control technologies have traditionally been applied to generation and transmission, the smart grid era is pushing sensors and, thus, visibility, into the distribution system, where the grid presumably “ends.” But the smart grid era is also an era in which distributed resources – in both utilities’ and customers’ hands – are making the “end,” or grid’s “edge” much more difficult to define.
Though we cannot offer definitive answers to the challenges posed by Big Data emanating from the smart grid, we would like to generate discussion of this multidimensional challenge. Big Data touches upon notions of ownership and stewardship of data and the rights and responsibilities of various stakeholders, including the security and privacy of their data. It touches upon the business case for data analytics. It calls for a debate over the relative merits of centralized versus distributed processing or a hybrid of the two. And, with the advent of the Internet of Things (IoT) – which makes Big Data and smart grids look manageable – it’s about a brave new world of Big Data at previously unimagined scale.
Same Old IT Challenges, At scale
We would suggest that traditional IT practices remain useful. Data needs to be relevant, clean, organized, aggregated, processed, and analyzed to obtain and present actionable intelligence. One challenge in the smart grid example is scale. Consider a likely case-in-point. Say a utility has 2.2 million customers and, thus, 2.2 million smart meters generating 3.3 terabits of data every day. In comparison, its SCADA system, serving about 300 distribution substations, produces 65 gigabits of data per day, only 1/50th the amount of data from smart meters. The amount of data from smart meters in one year is 1.2 x 1015 bits, or 150 Terabytes. Retrieving, storing, analyzing, and presenting insights from data on this scale is a daunting task.
Two related ICT issues also call for aggressive, consistent efforts, if not resolution: cybersecurity and energy use. As previously unimaginable numbers of devices are connected via the IoT, potential attack vectors also increase. Cybersecurity measures will have to be designed into devices and networks from the start. And new, markedly elevated levels of data collection, network traffic, and processing will place new burdens on efforts to make ICT sustainable, which currently consumes increasing amounts of energy and, therefore, threatens to increase levels of carbon emissions.
The Business Case
The scale of Big Data is daunting but utilities will deal with it for a very good reason: money invested in data analytics has a positive return on investment (ROI), estimated to be as much as $11 back for every $1 invested. Data analytics will drive big improvements in grid-related safety, reliability, resiliency, and operational efficiencies, as well as driving resource planning, capital investment, business decisions, and customer-facing programs.
Data-Related Stakeholders
With advanced near real-time data analytics, anomalies, trends, possible security breaches, and other costly business interruptions can be detected early and profit-generating changes can be made. The cost, investment and resulting value of Big Data and data-driven insights have different impacts on the grid’s various major stakeholders, which include producers/generators, transmission and distribution entities, consumers, and regulators. Each of these stakeholders has different roles and responsibilities in managing, storing, processing, protecting, owning, and using data.
Producers/generators use data analytics to accurately forecast demand, predict peak power needs for residential as well as commercial consumers and establish demand-side management programs. Transmission and distribution entities use analytics to identify anomalies in power delivery, detect, and avert outages before they happen, and restore service faster after an outage. Consumers can reduce costs by setting power-heavy machines such as dishwashers, clothes washers and dryers to run in off-peak hours. Consumers will demand value in return for the use of their data and it increasingly appears that data itself will become a value stream whose costs and benefits are apportioned to various stakeholders. Regulators use data analytics to ensure regulatory compliance from the point of power generation all the way to the point of consumption as well as to analyze the costs and benefits to consumers and utilities.
Strategies Needed
As sensors proliferate and become integrated into virtually every piece of equipment on the grid, it isn’t logical, feasible or cost-effective to return the entire tsunami of data upstream for centralized processing. In some cases, sensors are being integrated with processors so that only significant data or processed data need be communicated upstream.
In other cases, what has come to be known as fog computing can establish a direct connection between any individual sensor and the cloud so that rather than taking data to the analytics, one takes the analytics to the sensors generating the data. Yet the need to truly understand what’s occurring all across the grid may limit the usefulness of such distributed processing. In certain circumstances or at key times, fully granular data is needed to make automated or human-enacted decisions.
Debate continues on how much data a utility is entitled to have concerning customers’ energy use and customer-side distributed generation. The grid’s “edge,” as noted earlier, is becoming a more diffuse concept with each passing day. And the IoT’s promise to connect innumerable devices will only add urgency and scale to the challenge of harnessing Big Data.
We suggest that this challenge is urgent. If we can manage it properly, we will find a ROI-positive path to solving operational and business issues.
