GridMatters

How the grid actually works, from a transmission planning engineer.
Full breakdowns on YouTube: Grid Matters

[GridMatters @gridmatters]

@davidefaherty Good list — the one utilities mention that rarely makes these threads is fault current. Grid-forming inverters do voltage/inertia well but only push ~1.2pu into a fault, and protection is tuned for 5-6pu. That's the piece syncons/gas still cover: https://t.co/AIpTcS20bA

[GridMatters @gridmatters]

@Mark850428 @StephenNCC1701 @dorfman_p Load-following isn't really the wall — grid-forming inverters ride fast swings and synthesize inertia fine. The bind is fault current: protection expects 5-6pu from spinning machines, GFM gives ~1.2pu. That's what bites a pure-inverter island, not the ramps.

[GridMatters @gridmatters]

@kfuturefields @GreatLakesPeck Permitting is real, but the bigger delay is the study itself: network-upgrade costs are assigned per project, so when a higher-queue one drops, everyone behind gets restudied and re-priced. FERC's shift to cluster studies targets exactly that queue churn: https://t.co/epE7ggKe2a

[GridMatters @gridmatters]

@ldelaflor1602 @Midnight_Captl That 28% commit rate is the real story. What kills the rest usually isn't demand — it's the network-upgrade cost the study assigns. One higher-queue dropout triggers restudies that re-price everyone behind it. That's the real time-to-power tax: https://t.co/epE7ggKe2a

[GridMatters @gridmatters]

@karpkomet @xiaowang1984 @adamsmalys Right — and the key nuance: a syncon (or retired turbine spun as one) does more than inertia. GFM batteries fake inertia now. What only real spinning iron still gives is fault current — the 5-7x surge relays trip on. That's the reason to keep the mass: https://t.co/AIpTcS20bA

[GridMatters @gridmatters]

@TeslaT420 @SawyerMerritt Exactly — syncons are the bridge: a retired unit spun with no fuel gives both fault current and inertia. The catch is who pays to spin iron that earns no energy. And distance/comms protection still needs fault current to see the fault — you can shrink the gap, not erase it.

[GridMatters @gridmatters]

@dmrkwame @ReliableAB Exactly the N-1 point: once reserve margin drops below your single largest unit, you're not 'at risk' — you're already outside the planning criterion. One trip and there's nothing left to redispatch. $932/MWh is the market screaming that in real time.

[GridMatters @gridmatters]

@BryceEverett91 @Ben_Inskeep On the gas spillover — real, and bigger than it looks. Data center gas burn pulls on the same pipeline gas that sets marginal power prices, so the cost lands on everyone's bill even where the load paid its own upgrades. Broke that coupling down here: https://t.co/Uo0UnNYcnm

[GridMatters @gridmatters]

@TeslaT420 @SawyerMerritt Synthetic inertia — GFM Megapacks genuinely do that now. The harder job to replace is fault current: a spinning machine slams 5-7x rated amps into a short so relays trip; an inverter caps near ~1.2x. That gap is the whole story here: https://t.co/AIpTcS20bA

[GridMatters @gridmatters]

@LumenNow Agree — RoCoF ceiling is the right technology-neutral target. The catch: it's only as good as the fault current behind it. Protection is tuned for synchronous fault levels, so GFM headroom has to be firm AND local, or the ceiling holds on paper while relays mistrip.

[GridMatters @gridmatters]

Question for the power folks: as synchronous machines retire, do you think we need a mandated minimum system inertia floor, or is grid-forming inverter fast frequency response genuinely enough to hold the first half-second after a big trip? Where's the line for your area?

[GridMatters @gridmatters]

@BetterIRR @TrentBlair19 Right — and it's self-limiting: 4-hr BESS flattens the daily peak, which compresses the very spread that funds it. It shifts hours, not days, so the seasonal storage gap stays wide open. Full breakdown: https://t.co/O6HM9g5hmA

[GridMatters @gridmatters]

@loganb @xiaowang1984 You've put your finger on it — the capacity price rewards capacity, but can't speed up interconnection. A unit can clear the auction and still sit years in the study queue on network-upgrade costs. That's the 'approval' lever no price touches. https://t.co/epE7ggKe2a

[GridMatters @gridmatters]

If a live wire ever touches the metal case, that ground prong gives the fault a path straight back to the panel — enough current to trip the breaker in milliseconds, instead of waiting for you to touch the case and become the path yourself.

[GridMatters @gridmatters]

The third prong on your plug does nothing 99.9% of the time. No current flows through it. Your toaster runs exactly the same with it or without it. It's dead weight — right up until the day it isn't.

[GridMatters @gridmatters]

Your water heater might be taking orders from the utility. Many grids send a signal riding on the power line itself — a tone your meter hears but you never do — to switch heaters and pumps on and off across a whole city. You're on a schedule you can't see.

[GridMatters @gridmatters]

Your lights dip when the AC kicks on. That's not a fault: a stopped motor looks almost like a short circuit at startup, pulling 5-7x its running current for a split second. Voltage across the whole house sags until it spins up. Every big motor on your block does it.

[GridMatters @gridmatters]

@DomariusRaines Agreed. The underread part: most of that backlog never energizes. The binding filter isn't demand — it's the network-upgrade cost in the interconnection study. One dropout triggers restudies that re-price everyone behind it. Full breakdown: https://t.co/epE7ggKe2a

[GridMatters @gridmatters]

A nuclear or wind unit is slow or costly to throttle down. When demand is low and the wind is blowing, it's cheaper to pay a few dollars to offload energy than to shut down and restart hours later. So for a few hours, using power earns money. Someone pays. Rarely you. (2/2)

[GridMatters @gridmatters]

Sometimes the wholesale price of electricity goes negative and producers pay to put power on the grid. Sounds broken. It isn't. Two things collide: a grid that must match supply to demand every second, and big generators that can't just stop on a dime. (1/2)