The US Department of Energy (DOE) is readying the power grids for solar power to meet 100%, and greater, of local electricity demand as solar peaks mid-day. The DOE is funding a series of projects whose broad goal is to make the power grid more able to be monitored and managed while dealing with less predictable and controllable forms of electricity generation.

If you’re considering solar, get a quote from multiple contractors at understandsolar.com. If you want feedback on the quote you get – either email me at john @ 9to5mac dot com or send a tweet.

On May 19th, the DOE held a webinar (slides) titled, “Enabling Extreme Real‐time Grid Integration of Solar Energy (ENERGISE).” The DOE that solar power is growing exponentially and prices are falling quickly, and both of these patterns are expected to continue. By 2020, the DOE wants a grid that is a “highly scalable distribution system planning and real‐time operation solution that seamlessly interconnects and integrates high penetration (>50% of distribution peak load) solar generation.”

2020 is less than four years away.

By 2030, the DOE seeks, “transformative and highly scalable technologies that…enable extremely high penetration (> 100% of distribution peak load) solar generation.”

The DOE plans are summed in one of their slides:

sunshot

The first goal of the DOE is managing the flow of variable energy sources – the sun rises and sets and flows in a different direction on the power grid. The DOE wants to do this managing of the flow via smarter solar modules that report their production, smarter solar inverters that manage their output, smarter buildings that report their usage, better hardware that programs when it needs energy, and real time reactions by the traditional power company to tap needed resources filling in gaps. The power companies will be given the data to fill those by an ‘Enhanced System Layer’ of software that is tying together the aforementioned layers of users and producers in real time.

highlevel

 

The high level system requirements show that the level of intelligence on the current power grid is limited to a small number of central nodes. An increase from 10s of nodes on any grid to “~1,000,000” will mean significant deployments of hardware in small distributed packages at transformers, solar power systems, commercial and residential structures, all that will communicate their present state to a central optimizing database while waiting for directions if a change is needed.

requirements

One could say that the US electricity grid is about to wake up. Already, large countries are managing high levels of renewables,  Some private companies are already trying to smarten up the grid – AT&T & SunPower recently started tying together communications with module level data, and Enphase is attempting to make Hawaii’s solar industry react in real time. Microgrids and grid level batteries will also benefit from grid intelligence like this. With all of this, we see that solar and energy storage growth is accelerating and that the great driver of all of this is the falling price of solar power…maybe, just maybe, Kurzweil will be right.

A great technical write up, and the place where I first read on this project can be found UtilityDive. All slides came from DOE presentation referenced in the article and located here.