Two series and a handful of standalone essays on grid reliability, SRE for utilities, BTM economics, and what it actually takes to modernize infrastructure you can't turn off. Plain text, specific numbers, no transformation decks.
Seven-part series porting the Google SRE methodology onto distribution reliability. Budget, don't scorecard.
Network engineers have known for two decades that reliability is a budget, not a goal. The same math works on a 12.47 kV distribution feeder, if you stop treating SAIDI as a scorecard and start treating it as a currency.
FLISR is network failover. Reclosers are retry-with-backoff. Storm drills are chaos engineering. You just don't call them that, and you don't do them systematically.
Contingency planning gives binary pass/fail results for a probabilistic world. The 2003 blackout proved every component can pass and the system still fails.
Utilities already practice chaos engineering. Storm drills, black start exercises, protection relay testing. The missing piece is making it systematic, continuous, and measured.
How SRE practices complement IEEE 1366 reliability reporting. Per-feeder error budgets, burn-rate tracking, and real-time SAIDI forecasting layered onto existing regulatory frameworks.
A 30-minute SAIDI improvement avoids $10 million per year in outage costs. A typical SRE program costs $2 million. The math works at every scale.
SRE adoption at a utility is an 18-month organizational transformation. The cultural barriers are real and solvable. Here is the roadmap, from existing frameworks to new roles.
Four-part series on deploying AI assistants at utilities without leaking CEII, breaking NERC CIP, or waking up legal counsel.
Your utility engineers are already using AI tools, without governance. Early adoption disasters show what happens when AI runs ahead of policy.
What CEII, NERC CIP, and BESS regulations actually require for AI tools at utilities, and how the data tiers of modern AI assistants map to compliance.
A three-zone reference architecture for deploying AI coding assistants at utilities, with network diagram, data classification matrix, and NERC CIP compliance mapping.
A 90-day roadmap for deploying governed AI coding assistants at a utility, from data classification through pilot to full-scale rollout.
Standalone writing on modernization, iteration, and the operator's view of the grid.
Autonomy at the grid edge: a five-rung ladder from offline reasoning to coordinated autonomous substations, the open-source stack (Gemma 4 E4B, Hermes, MCP) that makes it possible, and the real-time OT use cases (VVO, DER dispatch, adaptive protection) already within reach.
Why the pursuit of a finished grid is a trap, and how to find satisfaction in continuous modernization. Reframing infrastructure progress in an era of perpetual evolution.
Why starting over makes you better, and how embracing iteration leads to mastery in software development.
Why iteration beats revolution in grid transformation, and how utilities build resilient infrastructure through persistent, incremental upgrades.
How utility leaders foster a culture of continuous adaptation in the face of decarbonization, electrification, and grid evolution.
A practical architecture for deploying AI coding assistants at utilities while protecting CEII, NERC CIP assets, and BESS operational data. The risks are manageable; here is how.