A futuristic grid driven by DERs will require a structural shift from traditional hub-and-spoke systems to fully distributed networks. What does this mean in practice and why is network structure important for the energy transition?
There are three basic designs of distribution systems – radial, looped, and networked – determined by topology: how the poles, wires, and substations crisscross a given service territory. These terms define a gradient describing the grid from most simplistic and least fault tolerant (radial) with many spokes protruding from few central hubs, to most complex and fault tolerant (networked) with many interlocking loops and spokes per hub distributed more evenly throughout the network. Because networked systems require building more infrastructure, they are expensive and predominantly found in population centers where the costs of added redundancy can be spread over many consumers. Cheaper radial systems cover the vast, sparsely populated areas of the country; measured in total line miles, the majority of the grid is technically not a network.
In addition to geography, network structure also depends on scale or granularity as measured by the power flowing through wires and their proximity to the final point of consumption. As shown in the visualization above, more simplistic networks tend to emerge at lower voltages closer to where power is distributed to consumers, or in other words, at high granularity (see the little houses on the left). Networked grids with high fault tolerance tend to appear at larger scales, on the transmission systems of bulk grids where it is cost effective to ensure that entire neighborhoods, towns, or regions have multiple avenues to electricity in the event a line goes down.
“Fractal” is a useful term to describe these grid characteristics because it offers a formal way to analyze network structure at different granularities. Originally coined by Benoit Mandelbrot, fractals refer to patterns that exhibit self-similarity or recursiveness (i.e. looks roughly the same) at different scales. They are properties of many different disparate naturally-occurring phenomena, from coastlines to broccoli (notice how the structure of a tiny broccoli floret mirrors the shape of the entire stalk). Additionally, fractal systems are typically associated with resiliency due to their self-similar structure and ability to behave independently at every level of granularity.
The network structure and technologies of today’s electricity grid exhibit limited fractal-ness. Some combination of a highly networked grid with renewables, storage, and gas generators appears to be a winning formula for delivering cost-effective resilience at both the ISO and microgrid levels. But as we move up the grid from resilient systems within homes, business, and campuses to the control room of an ISO, we see a change in structure and functionality at the intermediate distribution levels; they are predominantly radial and incapable of operating independently.
So investments in the hallmarks of structural resilience appear at the largest and smallest scales of the grid but they are not truly self-similar fractals; to be fractal, resiliency would be distributed evenly throughout all hierarchies to optimize not just for uptime on the bulk system and end nodes, but at the intermediate layers constituting the neighborhood or community level as well. Distribution networks serving these communities rarely, if ever, have the capability to island. By implication, resiliency is an all or nothing proposition for most of the population: you either pay for a backup system or are completely beholden to the bulk grid.
Do we need a truly fractal grid? If we can power homes and critical businesses when the transmission system goes down, why should we target resiliency within entire distribution networks?
This SunRun whitepaper, published in 2021, proposes distribution-level resiliency solutions for communities affected by wildfire-induced public safety power shut offs in California and paints a compelling picture of the benefits provided by a truly fractal grid. SunRun envisions a future where enough generation capacity and supporting hardware has been deployed on distribution networks to operate independently as needed, providing community resilience during service interruptions and load shaping services during normal conditions. This futuristic network is intelligent, in that it can smoothly transition between interfacing with the transmission system and operating independently as a snack-size ISO, coordinating the services of a diverse portfolio of privately-owned DERs through market constructs.
Several fruitful themes are packed into this vision. By operating at the neighborhood level, even those without the financial resources or technical means to install an on-site DER can access resilience when the bulk grid goes out; the community solar plant can still provide power to local customers, and neighbors with backup systems can still share or trade power. This creates a more equitable distribution of resilience while maintaining private ownership of assets through a method of centralized coordination by the Distribution System Operator (DSO) and decentralized control by community stakeholders. It also provides a community with the optionality to exit the transmission system if, or when, it is in its interest.
What is also captivating is that the robust system awareness required to run this community microgrid maximizes the value of distribution grid infrastructure. As we electrify heating and transportation, squeezing every watt out of increasingly strained wires, substations, and pole-top transformers will be essential; the key question is whether the telemetry and software deployment necessary to provide this degree of flexibility is cost-prohibitive. However, as grids continue to see delivery charges rise as a proportion of total costs, investing in these sophisticated islanding capabilities may soon become economically competitive with the status quo – especially in areas with frequent service interruptions.
Continued in Part II.
The idea of local resilience and the fractal grid is indeed alluring. However, the costs and technical/safety challenges of building this are prohibitive with today's technology.
If a wire falls to the ground, there needs to be a way to isolate both sides of the fault before the supply can be restored. At the low-voltage grid level, this technology is not common and remains expensive. This sort of design would require devices that can sense a fault and isolate the area - deployed across the millions of miles/km of the LV network.
Once the fault is isolated, we need to have a system that can then balance supply and demand in real time and maintain the system frequency. DER is not able to do this today, but moves towards small-scale grid-forming inverters may make this more achievable. However, given the random nature of where a fault could occur, it would be difficult to have sufficient local generation (day and night) to always match demand - even if that demand was somewhat flexible.
Islandable communities could be economic where the community is relatively co-located and served by a long supply line.
For DER, the current economics (and physics) would suggest the islandable homes may represent a more cost-effective outcome that building a self-healing islandable grid.