Hacker News Reader

Hacker News

Comment by probablypower | Hacker News Reader

Parent Full thread

probablypower

This is an important point.

Increasing renewable penetration into a power system leads to:

-> reduced inertia

-> increasing Rate of Change of Frequency (RoCoF)

-> reduced time to react

If you rely entirely on manual systems to react to frequency changes on a grid with high renewable energy penetration (e.g. the Irish grid), you'd really struggle to avoid daily blackouts without decommissioning a bunch of wind farms.

It is by using such automated control schemes that countries like Ireland are able to push the limit on wind power penetration.

If I was being truly cynical, I would think that a push towards mandating manual control (rather than mandating that it is available as a fallback) is actually a way to limit growth of the renewable energy sector.

View on HN

While a smart grid is incredibly useful to solve this, is it strictly necessary?

For example you could build an array of flywheels that charges when the frequency is above 50Hz and discharges when the frequency is below 50Hz.

No network connection necessary, and if you want to you can build it entirely analog. If all you really want is inertia and reaction time it could even be as simple as dumb three phase motors with weights.

I'd consult the relevant project drawdown entry https://www.drawdown.org/solutions/electricity-generation/gr...

> For example you could build an array of flywheels that charges when the frequency is above 50Hz and discharges when the frequency is below 50Hz.

All rotary field machines (power plant generators, phase shifters, motors) work this way and contribute stability to the grid.

These basically add inertia (i.e. dampen frequency changes), but they do not regulate frequency. That's a fundamental difference.

A simple motor or generator simply add inertia, but I see no reason why you can't use a flywheel to deliberatly regulate frequency, for example by adding a continuously variable transmission.

But really inertia is all you need because all we are trying to do here is replacing the lost inertia from replacing spinning generators with solar. The recovered inertia gives human operators the time to make phone calls to increase energy production, spinning up a pumped-hydro plant or whatever is available.

Who is going to pay to build and operate this flywheel inertia plant that generates no power? Sure it is possible but if you are going to spend that kind of money you can do a lot better.

The same people that would alternatively pay for smart grids: grid operators tha want to prevent blackouts in a changing environment. It's just another piece of infrastructre next to power lines and countless substations and transformers.

It's been done already! Beacon Power operates a 20 MW / 5 MWh flywheel plant in Stephenstown, NY that does nothing but store grid energy and release it when needed. And ancillary services such as spinning reserve carry a price on the electricity market, so there is definitely an opportunity there.

I'm not sure if frequency regulation is currently considered an ancillary service in electricity markets right now though - from what I know about NY state, only 10- and 30-minute reserve are priced.

Usually power plant generators are required to respond to system frequency changes with a 5% droop characteristic so that they do regulate frequency by arresting dips and spikes.

We can do analog control. Almost anything digital can do, analog can as well... although often with very complicated implementations.

This.

As someone responsible for the IT systems reliability of a grid operator I have to wonder if this really is a backhanded attempt to limit to the growth of renewables.

We run load frequency control on the grid every 4 seconds to keep the grid in balance. Hard to see how that'd work with manual controls on today's complex grid.

Or, nuclear.

does modern nuclear eschew digital controls? i mean... i’m sure they have manual cutoffs and sensors that humans monitor; can’t have some rogue hacker able to turn a nuclear plant into a huge catastrophe from the other side of the world and all...

but i’d think they’d be largely automated to reduce human error?

Parent was talking about the necessity of networked digital control systems to manage intermittent renewable sources.

My point was that a grid comprised of a higher percentage of nuclear wouldn't need as advanced networked feedback, as sources wouldn't be turning on and off suddenly.

Indeed there would be local digital monitoring and control with nuclear, but that's a much less risky threat vector.

Digital elements in control is not the same as internet controlled.

There is no valid reason to make nuclear power plant controls accessible for control by internet.

It would be crazy to make it network-controllable, but you could have a data diode to allow power generation metrics and other monitoring data pass from the control network to the outside world.

If we didn't let scaremongerers ruin nuclear with politics we could have had a 100% carbon-neutral (I don't count transportation because we can move to electric vehicles) grid with more cheap power available than we could find a use for by now.

Nuclear is statistically the safest form of energy in terms of deaths/kWh and waste storage is massively overblown, not even accounting for how we could be recycling that waste in new plant designs if it weren't for the fubar politics around it.

> Nuclear is statistically the safest form of energy in terms of deaths/kWh

This is somewhat misleading, because the major risks are things like birth defects and cancers which are really hard to attribute to any specific cause. Not to mention that it's stochastic; we probably won't have a disaster, but if we do it ruins a huge land area at tremendous cost.

Also, this doesn't work for every country.

Exactly, according to Russia, only 31 people died as a result of Chernobyl.

Reply on news.ycombinator.com