Distributed intelligence for a DER-based grid
Intelligence in smart meters and other IoT edge compute network devices, whereby applications utilize analytics to provide near real-time situational awareness and localized control, is set to play a key role in the operation and ever-increasing adoption of renewables and electric vehicles, is discussed by Itron’s Wassim Akhdar and Nick Phillips. This insight and control provides significant value on top of the operational benefits gained from an AMI network.
Utilities need to manage their existing infrastructure more efficiently with better information while planning their grid modernization strategies.
In the US, China and Europe, the electrification of transportation is destined to create significant loads on their low-voltage circuits. The global movement to decarbonize energy sources is going to significantly change the structure of electricity generation and how the grids are designed and maintained.
This will require moving the ability to monitor and manage load to the grid edge. Ensuring a reliable supply of power in electric distribution systems will require discipline at the edge in maintaining proper voltage, power flow and frequency in the medium-voltage distribution and low-voltage circuits.
“It’s all about the quality and fidelity of the data that produces actionable information and drives operational and planning improvements.,” says Nick Phillips, Senior Manager of Technical Sales in Asia Pacific for at Itron, which has pioneered the concept of distributed intelligence (DI) in the utility sector.
“Distributing the intelligence can reduce the latency between making decisions and taking actions using significantly higher fidelity data that produces precise results, such as a more resilient grid, insight into operations, improved customer engagement and increased safety.”
As an example, he cites the emerging application of electric vehicle (fleet) charging optimization, which is aimed at operators such as bus and taxi companies with large fleets of electric vehicles and has attracted some early interest in India.
“With a large fleet of electric buses, making sure they are electrified and ready to use would put huge pressure on the infrastructure. If every charger was able to determine which chargers around it were on and what the stress on the local network was and switch on or off in a coordinated way, then one doesn’t need to collate that into a large central data processing infrastructure to do those calculations.”
The distributed intelligence value proposition
The electric vehicle (fleet) charging optimization use case is reflective of the wider challenge facing distribution grid operators with the rapid growth of DERs and the complexity of the ensuing power flows.
With that growth is the accompanying growth in the volume of data becoming available, which is being further accelerated as the data generation frequency increases from the previously common hourly or half-hourly meter reads down to the five-minute or even shorter intervals.
“There is a need for visibility and control at the grid edge because of the evolving sophistication of how consumers are using electricity, and that’s lacking in the systems that exist today,” Itron’s Director of Product Management Wassim Akhdar, comments.
“While we can now pass larger amounts of data through communications networks, we are trying to solve problems that are highly complex; real-time streaming of data through to the back-office isn’t feasible, and in some cases, the computational complexity and real-time response makes it too difficult to do in a centralized location.”
DI provides critical information for planning and maintenance, thus reducing operational costs. DI can move utilities from a “run-to-fail” model to a “reliability-centric” maintenance model that improves reliability and sustainability. DI uncovers how your grid is actually responding to loads at the edge of the grid versus relying on assumptions made by engineering models.
In this context, again an example is electric vehicle (EV) charging, with a set of domestic chargers in a street required not only to not overload the local transformer but also take into account the users’ individual vehicle availability requirements, with such an exercise potentially more efficiently solved computationally within the meter than in the utility back-end office.
A growing ecosystem
Along with the availability of DI, a growing set of use cases have emerged which have been developed into ‘apps’ by Itron and a growing ecosystem of partners and are available for download from an enterprise application centre to a set (or selected subset) of DI-enabled devices.
These are currently focused on the management of the distribution grid with use cases including active transformer load monitoring, bellwether voltage monitoring, high impedance detection, meter bypass detection, and location awareness.
Others address DER integration with active premise load shedding, solar and EV awareness, consumer engagement with data cloud services and smart payment use cases.
“As one brings in more DERs, real-time monitoring is needed to maintain the reliability and quality of supply,” says Akhdar, noting how a meter or group of meters can communicate with one another to take action and for example, stop customers with solar photovoltaic generation from pushing excess power to the grid.
“This is a classic reliability issue and one of a number of examples in which utilities can really get the most out of the network in terms of energy efficiency.”
At the opposite end of the grid reliability spectrum is basic security of supply and Phillips points to countries in Asia Pacific with less advanced economies in which AMI is new and they are aiming to deliver reliable supply to drive economic growth.
“For some people, just having any power is important. Being able to respond and get power back after an outage due to a severe weather event is helpful if one can understand where the fault occurred, down to the level of between one or two connection points as this is important for the country’s economic growth.”
Valuing the use case
Itron’s experience with cost-benefit analyses of app implementations, which are undertaken before development or implementation, shows the benefits are generally upwards of 2:1 and closer to 3:1 if the benefits can be stacked.
This compares with the approximate 1:1 of an average AMI rollout.
“It obviously varies between countries and regions and even among utilities, depending on their specific needs, and that’s a real distinction with the DI apps as they can be targeted to the specific use cases,” comments Akhdar, adding that the cost-benefits have been confirmed in the field.
Additionally, there are potentially unanticipated benefits, with one example being when a North American utility using the High Impedance Detection app detected a case. The utility planned the dispatch team to replace the service but instead, they traced the fault to a squeezed connector on the transformer.
“In that case, the avoided costs were that of replacing the service plus the outage time the customer would have endured during the replacement, so DI saved the utility quite a bit of money, reduced the cost of the dispatch, and increased customer satisfaction.”
Distributed vs. back-office
With over 11 million Itron DI-enabled meters contracted and shipped, and over 1 million running apps, DI adoption is scaling rapidly.
The nature and urgency of the use case will determine whether back-office analytics or edge-base intelligence makes the most sense to utilize. “Those use cases where we need to physically process the data will be done so in the back office,” says Phillips mentioning as an example the traditional use of AMI data for billing.
“Where we need to see derived data only or need to take action very quickly would be a classic distributed intelligence case, whereas if we need to record and retain the data for regulatory or consumer activity, it needs to come to the backoffice.”
There are also some use cases with both DI and back-office applications, such as meter bypass detection (DI) and revenue assurance (back office). In addition, solar and EV awareness also can be undertaken in both places, with data granularity and latency as key determinants of the optimum approach.
DI looking ahead
Looking ahead, both Akhdar and Phillips believe that an unexplored area with potential for DI is consumer and behind-the-meter use cases.
“With the existing meter—and without having to install any additional hardware at the customer side—it provides behind-the-meter analytics and applications such as load/appliance disaggregation and pre-pay functionality,” says Akhdar.
Phillips concurs, saying that apps are likely to emerge that haven’t been imagined so far, such as linking consumer behaviour and health through load monitoring, e.g. for an aged relative.
“For me the exciting thing is being able to solve new problems. We’ve gained more and more visibility into the network over the years and now with DI we are able to go behind the meter and detect bad connections and safety issues—these are things that matter.”
For more information visit:
https://www.itron.com/na/solutions/what-we-enable/analytics/distributed-intelligence
https://www.itron.com/na/solutions/what-we-enable/analytics/distributed-intelligence/value-and-applications
For more content from Itron, visit here
