As the deployment of distributed generation (DG) increases, electricity distribution utilities face challenges around integrating this new capacity into their grids. Once the existing hosting capacity has been depleted, this has usually demanded heavy capital investment in reinforcing the grid infrastructure to meet the emerging needs.
However, experience shows that traditional business-as-usual grid investments potentially create future inflexibility as regulatory and market needs change, but are also often higher need be. A major reason for this overspend is distribution businesses’ lack of sufficient visibility into how much, when and where on the grid DG will come onstream.
The underlying issue is that DG penetration doesn’t happen equally across any power grid. Instead, it tends to be concentrated in particular areas, as dynamics such as the “network effect” kick in—with customers in a particular neighborhood installing DG and then telling friends and neighbours locally how much they’re saving on their bills.
The result? A small number of substations on the grid experience especially high DG penetration, eating up their free hosting capacity and turning them into bottlenecks with an increasing risk of power quality issues. Currently, inaccurate DG deployment forecasts mean distribution businesses find it difficult to anticipate these pinch points, which in turn drive most of their reinforcement needs. This mismatch risks making a distribution business’s capital investment on grid reinforcement both inefficient and disproportionately high.
This is a challenge across the distribution utilities sector in all territories where DG is rolling out—which today is virtually everywhere. And it’s something most distribution businesses are aware of, reflected by their growing recognition that investing in most effective DG forecasting will result in getting more “bang for their buck” from their grid reinforcement spend.
Based on simple assumptions, the relationship between improved DG forecasting and reduced grid reinforcement capex would be roughly 1-to-1; A 10 percent uplift in the accuracy of DG forecasting could generate a 10 percent reduction in the related reinforcement capex.
However, in light of the wave of DG deployment across grids globally and the significant local impacts on hosting capacity, we decided to undertake research-based geo-modeling as part of our Digitally-Enabled Grid (DEG) research program, to answer a key question: How much value do improved DG deployment forecasts deliver for distribution businesses in selected countries or states—and what are the factors driving this value?”
To answer this, our value modeling drew on country and state-level data and academic papers from five markets worldwide—the United Kingdom, the Netherlands, Australia, California and New York. We supplemented this with more than 9,000 sets of substation-level data from between two and five leading distribution businesses in each of the selected markets, together with reviews with nine experts from local transmission and distribution teams and an external grid sector analyst. Finally, we designed the modeling to test the results’ sensitivity to different assumptions and market scenarios, with a view to identifying the key reasons behind different outcomes for different geographic locations.
What did our modeling tell us? Essentially that DG deployment typically results in an amplified impact on hosting capacity needs, driven by the unequal deployment of DG across regions and substations. The unequal spread means a small proportion of substations become “hot clusters” taking most of the new DG coming on stream, against a background in which the “hot cluster” substations already have above-average reinforcement needs.
To put this in hard numbers, we found that a 10 percent forecast inaccuracy for DG deployment at the local level can lead to forecast inaccuracy of more than 20 percent for reinforcement spend. At first, this multiplier effect sounds like a challenge. But it’s only one side of the coin. The other side presents a major opportunity: a 10 percent improvement in forecast accuracy can result in a distribution business saving between 12 and 28 percent of grid reinforcement spend across the five markets in our study.
The message is clear: better DG forecasting can more than pay for itself through lower capex. So how can distribution businesses achieve more accurate forecasting? Key enablers include improving analytics capabilities, local grid modeling and distributed energy resources management systems (DERMS) solutions, and forming partnerships around data to glean insights from potential DG customers, historical trends and other markets. And the savings achieved can be invested in new solutions such as energy storage and demand response to further reduce traditional spend.
Thanks to Sankara Narayanan from Accenture Research for his contribution.