Kraken's Serverless Power Grid: A Technical Deep Dive

Serverless Power: A Technical Blueprint

The presentation by Kevin Bowman from Kraken Technology offers a compelling case study of using serverless architecture to manage the complexities of a renewable energy grid. The key insight is the utilization of AWS services like Lambda, EventBridge, and DynamoDB to handle real-time frequency response and predictive optimization for grid-scale batteries and virtual power plants. This is innovative because it demonstrates the scalability and agility of serverless for a traditionally complex and latency-sensitive industry. The article highlights the need for millisecond-level responsiveness, which is achieved through the serverless stack. However, the analysis lacks specific performance metrics (e.g., latency numbers, cost savings compared to traditional architectures). The reliance on AWS services also introduces vendor lock-in, which should be considered. Furthermore, while the presentation offers a high-level architectural overview, it doesn't dive into the nitty-gritty details of the algorithms used for predictive optimization, which could have provided more technical substance. Nonetheless, this is a very interesting application of serverless.

Key Points

• Kraken Technology uses a serverless architecture (AWS Lambda, EventBridge, DynamoDB) to manage grid-scale batteries and virtual power plants.
• The architecture addresses challenges like millisecond-latency frequency response and 24-hour predictive optimization for renewable energy integration.
• The solution focuses on smoothing out the peaks and troughs in renewable energy generation, like wind and solar, to ensure grid stability.
• The talk highlights the importance of battery telemetry, control systems, and an ‘autopilot’ algorithm for automated battery management within energy markets.

---

📖 Source: Presentation: Kraken's Serverless Architecture for Keeping the Grid Green