At Google I helped define a distributed computing approach which scaled us more than 10,000x. As the grid makes a similar jump, from 1000s of reliable generators to millions of unreliable participants, what can we learn?
In a blog post with ESIG* published last week, I shared learnings from my time at Google in the context of how grid operators might apply the techniques and tools of the internet giants to the challenge of managing millions of distributed energy resources. This post builds on thoughts I shared in May (blog here, tweet thread here).
Broadly speaking, I share three suggestions to help develop systems for managing millions of DERs:
- Reliability is a property of the system, not the components
- Hierarchical control provides a healthy balance of central coordination with decentralized control
- Comprehensive visibility is the first step that today’s grid operators must take to enable more dynamic system operation
Now one really large caveat, which I want to explicitly call out, is how much of this problem space *isn’t* technology driven. This is 100x the case for the electric grid, which has to handle issues like legacy system integration, equity & fairness, regulatory considerations, and business model concerns. Still, used correctly, software can act as a force multiplier for the people on the front lines of driving change. If we can increase the pace of system transition for the grid by leveraging existing software approaches, we’ll all move faster.
If you’re interested in learning more, I’d encourage you to read my ESIG guest blog and reach out to me on twitter (@shinynew_oz) or via email with thoughts or questions. And for an example of how we at Camus Energy are putting this learning to work, check out this article that includes a section on our Taos project.
*The Energy Systems Integration Group is a wonderful organization bringing together different stakeholders at the forefront of the energy transition. If you haven’t engaged with them already, I highly recommend doing so.