Analyzing How ARPA-E Is Spending $33 Million on Distributed Generation Control Software and Hardware

The U.S. Department of Energy’s advanced project agency, ARPA-E, has awarded $33 million to 12 organizations to develop better ways to control distributed generation assets. Each of the winners of the so-called NODES program is aiming to coordinate power load profiles and generation assets to proactively shape electric loads. ARPA-E describes the technology as “virtual energy storage,” in that this distributed generation control software and hardware approach may alleviate the need for costly batteries (client registration required). Analyzing the winners, we find that aside from two national laboratories, most of the winners are clustered either in the U.S. Southwest or Northeast, with a fair distribution in terms of recipient type.


University research groups captured most of the $33 million of funding, although per-awardee funding was significantly lower for them. On the other hand, the two corporations that did win funding – the established heavyweights General Electric and Eaton – enjoyed the highest funding awards, along with the national laboratories. Startlingly, no startups were recipients of NODES funding, in stark contrast to other ARPA-E programs in the past for other research topics.


In terms of technical approaches, all of the awardees will look at the problem of how to deal with the real prospect of more than 50% of power generation coming from intermittent renewables, particularly distributed solar, but also other technologies like wind turbines. Coordinating millions of distributed assets poses a tremendous software and hardware control challenge, and the NODES awardees’ approaches feature some common themes: Real-time control is viewed as practically a necessity, with half of the teams highlighting it outright in their proposals. To enable real-time operation, some teams are pursuing improved sensor protocols to estimate transmission and distribution network demands, while others are looking to implement weather information and even forecasts.


Cloud-based platforms are viewed as important too, with 25% of awardees pursuing this feature – but making control systems both real-time and cloud-based will not be an easy feat. Finally, a third of the proposals emphasized that their approach will be hierarchically structured along multiple tiers. However, exactly what those tiers are composed of differs: Some projects break down the problem by location (transmission, distribution, and behind-the-meter) while others break it down by time scales (faster class of options for device-level control and slower for network-level).

While most of the projects are not focused too much on applications at this stage, there was significant interest on enabling demand movement and frequency regulation in particular. As the control technologies develop, the list of applications targeted will undoubtedly expand in both scope and specificity. For now, clients should view this development as a positive boost for a crucial part of the distributed generation puzzle – improved control software and hardware, focused in particular on high renewable penetration rates. However, clients should be mindful that many of these projects are competing, not collaborating: The resulting platforms will likely be vying for market share against each other, so it will be crucial to track commercial partnerships and deployment traction. Moreover, the alarming lack of start-ups in NODES’s funding points to an opportunity to enter the space, or back nascent companies, or help one of the six universities awarded here spin out its efforts.