While the costs of distributed energy resources (DER) continue their remarkable decline, potentially more important new data reveals their growing value to both customers and the bulk power system, experts say.
Planning by utilities and system operators has traditionally been based on complex forecasts about bulk system factors like resource cost and customer demand to balance supply and load. Local resources have been an afterthought. A new modeling tool claiming unprecedented processing power and detail offers the potential for a new level of joint transmission-distribution planning strategy not undertaken before.
"I don't know how this modeling tool fits into regulatory proceedings, but it's unreasonable for regulators whose duty is to protect customer interests to not investigate it," said former Maryland utility commissioner and Sunrun Chief Policy Officer Anne Hoskins. "If I were still a regulator, I would ask our energy staff to determine whether the commission should require its study or use."
Technical documentation of Vibrant Clean Energy's WIS:dom-P modeling tool shows it offers significantly more granularity about where and when distributed and other energy resources are available than other common planning tools. And more detailed planning can lead to better investments on behalf of customers, stakeholders agreed.
"If I ran a utility, I would definitely want to understand the added value of distributed energy resources because they will be an important part of the overall energy supply," Wells Fargo Renewable Energy & Environmental Finance head Philip Hopkins said. Knowing the value that DER offer the distribution system "will allow them to be used sooner and more efficiently."
But the new modeling tool faces two hurdles to overcoming doubts by traditionally-minded planners.
The biggest is how to get utilities and their regulators to compare it with current modeling tools. It must also demonstrate real-world validity of the computations, as Vibrant sought to do with "Roadmap for the Lowest Cost Grid," released Dec. 1. The roadmap shows hypothetically how computing power and granularity — in assessing where utility-scale and distributed resources will be most cost-effective — can find hundreds of millions of dollars in savings for utilities and their customers.
A better planning tool
In today's power system modeling, analytic software focuses on protecting reliability but emphasizes limited variables, like cost-effectiveness and resource capacity, over others. Planners understand the "uncertainty" in that modeling and are pursuing ways to "enable more-optimized resource plans," according to a 2018 Department of Energy overview of planning tools.
Vibrant's tool uses larger processing capability to optimize for more variables at a more granular level, according to Vibrant CEO and tool lead designer Christopher Clack, who previously co-developed National Oceanic and Atmospheric Administration modeling.
Unlike typical system planning, which focuses on five or ten years and no more than 20 years, Vibrant assessed U.S. system needs decades ahead in its hypothetical representations of generation, storage, transmission and DER, Clack said. And WIS:dom-P included policy factors, like renewables mandates, and engineering and system constraints in modeling system needs in 105,120 annual five-minute time segments at approximately 1 million 3-kilometer square, neighborhood-level sites, he added.
Regional Energy Deployment System (ReEDS) is the National Renewable Energy Laboratory's "flagship capacity planning model for the power sector," according to its website. But in contrast to WIS:dom-P, it models only 356 U.S. geographic regions and 17 annual time slices, according to the lab's 2019 documentation.
Aurora and PLEXOS, modeling tools from power system software provider Energy Exemplar frequently used by utilities for planning, are potentially configurable to greater granularity. But they typically forecast out only 20 years, with limited zonal resolution and hourly time slices.
The widely-known EnCompass tool from Anchor Power Solutions offers configurable geographic and temporal granularity that may match the Vibrant tool in granularity, but does not have WIS:dom-P's distribution-utility interface. The distribution-utility interface optimizes between bulk system resources and DER on the distribution system, Clack said. "I know it's a first-of-its-kind because I invented it during this project."
Planning for DER
The WIS:dom-P interface optimized between utility-scale and distributed resources in the hypothetical modeling and answered two basic questions, Vibrant reported. Its main scenario showed adding DER maintained "resource adequacy, reliability and resilience" at lower costs and met "clean electricity goals."
Vibrant first simulated U.S. energy supply and demand scenarios through 2050 using traditional modeling with and without the distribution-utility interface. It then ran scenarios with and without the interface, but with a national clean energy standard that would reduce carbon emissions 95% below 1990 levels by 2050.
The most cost-effective way to meet that 2050 clean energy standard is to deploy 247 GW of distributed solar and 160 GW of distributed energy storage, Vibrant concluded. The modeling showed it would save utility customers up to $473 billion cumulatively on electricity by 2050. Any transition to clean energy will lower customer costs, but using DER was $88 billion less expensive than transitioning with today's planning focused on utility-scale resources.
However, DER is not the whole solution.
Having more local DER available "unlocks the potential of utility-scale solar and wind," Vibrant found. "Retiring fossil-fueled power plants that run infrequently and deploying local storage more efficiently will help integrate 798 GW of utility-scale solar and 802 GW of utility-scale wind by 2050," Vibrant found.
The interface allows WIS:dom-P to recognize, unlike other modeling tools, that DER can "reshape" demand, the paper added. It can take advantage of accelerating customer investments in DER to obtain system load and peak load reductions and reduce spending on new utility-scale generation, transmission and distribution infrastructure.
This "reshaping" of demand also reduces peak period system ramping stresses, burdens on the transmission system, and enables the use of load flexibility on the demand side, Vibrant added.
With one planning tool that "accounts for total system costs and benefits," modeling will identify the solution with the "highest value," Vibrant said. It "will not select utility-scale resources with low energy cost if transmission and distribution costs to serve load are too high," but if those costs make utility-scale resources more cost-effective, the model "won't select more costly local solar."
Does it add up?
Not everyone is confident that scaling of DER will be affordable.
"Utilities and commissions should be looking for a better way to identify the value of local DER because they are now more important than ever," said Jeff Bodington, president of energy investment bank Bodington and Company. But "their characteristics are not completely known" and potential costs of reconfiguring utility system operations should not be overlooked.
More DER does create greater challenges, but it also adds opportunities for utilities, Wells Fargo's Hopkins said. "Solar on a home can contribute to midday overgeneration, but it also defers or delays expenditures for the distribution system, which can have a profound effect on utility costs."
The introduction of DER into operations can complicate the way a utility must manage reserves, he acknowledged. But if the utility has some control over the DER, it has additional flexibility, and "there will be savings opportunities in making that flexibility work."
Conventional wisdom is that large central station power plants are the most cost-effective way to meet demand and utility-scale renewables are falling in cost while local solar and storage are too expensive, said Rábago Energy Principal and former Texas electric utilities regulator Karl Rábago.
"But integrating and optimizing DER requires little investment, and sometimes small solutions can be more cost-effective than big ones," Rábago added. "Planners' cost estimates often overlook the costs to deliver central station fossil fuel generation and how scaling local solar and storage reduces overall system costs, allowing for more large-scale renewables to meet policy goals."
Distribution planning should not be separate from integrated resource planning (IRP), he said. "Plan the system with a full accounting of what's available and get the savings on the big stuff."
Utility system planners typically "solve for supply-side resources and then estimate how many distributed energy resources are cost-effective," Natural Resources Defense Council (NRDC) Climate and Clean Energy Program Senior Analyst Mohit Chhabra agreed. "Cleaving the analysis leads to imperfect estimates of the need for distributed resources."
But today's planning tools have long protected reliability and customer costs and will make regulators reluctant to consider new alternatives, some stakeholders said.
Will regulators try it?
No utility representatives were named as reviewers of the modeling, and state commissioners listed by Vibrant as reviewers declined to comment for this story. That may be due to potential conflicts in pending proceedings rather than a reluctance to endorse, but it suggests acceptance for WIS:dom-P could meet opposition.
There are still uncertainties associated with introducing DER into system operations, people interviewed for this story acknowledged. Availability and costs for technologies needed to integrated DER and to protect reliability are unclear, they agreed. And there is no guarantee customers will adopt DER at the pace needed to serve system needs, utility executives often argue.
"Regulators should consider this type of modeling, but there is lot of inertia in regulation of planning because the top priority is protecting reliability," Sunrun's Hoskins acknowledged.
In energy efficiency proceedings, Maryland regulators have required utilities to use alternative approaches to evaluating benefits "along with the planning tools they had always used," Hoskins said. "The utilities and regulators were not compelled to choose the alternative models' findings, but it is reasonable to think planning tools should change in response to changes on the system."
That change will accelerate as Federal Energy Regulatory Commission (FERC) Order 2222 is implemented, former FERC Chair and GridPolicy CEO Jon Wellinghoff said. The order requires tariffs and rules for bringing aggregated DER into wholesale markets.
When that happens, it will allow state commission-established distribution system market platforms, Wellinghoff said. Models like the one Vibrant designed "will be essential in market planning and operations because they show the value of DER to the whole system."
Another pathway for getting the new planning tool accepted in standard planning is for stakeholders to use Vibrant modeling in state proceeding testimony, he added. That could allow regulators and utilities to see both the value of the tool and the value of DER.
But there are "two daunting barriers" to including DER in planning proceedings, NRDC's Chhabra said. It is "computationally hard," and it is difficult and expensive to obtain "good quality data" on which modelers can make sound assumptions.
WIS:dom-P analysis "solves for both demand and supply side together," but it does not yet solve the location question that advanced distribution planning is working toward, he added. The next step is using its locational data to assure DER is built "in the right places."
Vibrant's tool "is exactly what is needed in state commission proceedings, in distribution planning and resource planning cases, and in the general policy conversation with legislators," said Environmental Law and Policy Center Senior Attorney Bradley Klein, a veteran of Midwestern state proceedings.
"The tools used to make critical regulatory decisions were built for a different set of resources and a different system and they undervalue local distributed resources," Klein said. "Decisions are only as good as the analyses used to reach them."
Vibrant's modeling tool "should be tested in litigated proceedings before commissioners," he said. "Experts can present the data and justify it, and highly motivated adverse parties can pick through the testimony to reveal its weaknesses. That will reveal its true value."
WIS:dom-P "can make cases stronger where commissioners are interested in data-driven evidence," Klein said. And "it has an important role in distribution planning, but it should not be confined there. We are considering using it in the upcoming Minnesota and Michigan IRP and distribution planning proceedings," he added.
Another way this can be useful is in the convergence between state retail markets and wholesale markets that will follow from FERC Order 2222, Klein said, agreeing with Wellinghoff. "That may be where it evolves over the next couple of years, as a way to quantify the types of services and compensations local resources on utility distribution systems should get in wholesale markets."
Decisions on "billion-dollar investments should not be made based on the output of a single model chosen by a single utility in an IRP proceeding," Rábago said. "The same old modeling leads to the same old solutions. This analysis asks better questions and gets better answers."
Update: This story has been updated to clarify comments from Sunrun Chief Policy Officer Anne Hoskins regarding Maryland's regulatory proceedings.