Hacker News Reader

Comment by WalterBright | Hacker News Reader

The article doesn't mention a fundamental point. Change the price of electricity from moment to moment based on the supply. This will obviate much of the need for storage, since a lot of electricity usage is elastic (car battery charging, HVAC, etc.) and can be time-shifted.

View on HN

On the other hand, the change in price from moment to moment is an arbitrage opportunity for anyone with a grid connected battery. As battery prices keep going down, more existing generators will be motivated to invest in them, which will gradually create an efficient time shifting of energy, which will gradually erode the existing price volatility.

I think that is a strictly good thing though; it is a flywheel solution (almost literally!) because the more people buy in and work with that model, the more successful it becomes. In a sense you gamify the electrical grid to train people to be more responsible consumers. If at the end of it we are paying the same price but the grid is more stable, I'm ok with that!

I heard a talk just today on this by Dr Dan McGillivray (Exec. Director of the Centre for Urban Energy @ Ryerson University, Toronto) at AI Toronto [0] who mentioned two projects, utilizing a mix of central power generation, decentralized power generation (via solar panels, etc), and the blockchain for transactions:

* the currently running Brooklyn Microgrid [1]

* and Toronto Hydro is apparently currently assessing a similar one (the blockchain isn't mentioned in this summary, but Dr McGillivray mentioned its usage in his talk today) [2][3]

There is still so much to be worked out, I could hardly digest all of the details he spoke about. It does sound promising, though!

[0] http://www.ai-toronto.com/ [1] http://brooklynmicrogrid.com/ [2] https://www.sdtc.ca/en/portfolio/projects/transactive-energy... [3] http://awsassets.wwf.ca/downloads/wwf_dg_report_long_bookmar...

This assumes automatic adjustment of consumption on the consumer end. I doubt individual residential consumers will look at the time and the consumption and plug their appliances according to the network for a few cents. I do not know if industrial usage is enough to move the needle here. Maybe someone can chime in with some data?

If (some) people drive five miles to save $0.10/gallon on gasoline, they'll run the dishwasher when power is less expensive.

However, your question presupposes that it's only human behavior that can be influenced with price. Send a price signal and let technologists and entrepreneurs do what they do best...

Some people doing that is not enough to assume significant shifts in market forces will happen. The question is will enough people do that?

I agree, a price signal might send some people finding solutions, but I'm afraid that a lot of the solutions would cost more than the savings, so that would reduce the field of what is possible. And mostly, I wanted to underline that simply sending a price signal is sometimes not enough for certain markets and does not scale too well by itself.

Gas prices vary daily. This works very well at matching supply with demand. I don't see any particular reason to believe that there's something different about electricity.

If I had a smart dishwasher or a smart washing machine that could turn on or prep hot water when the demand is low, it could work.

It could, but for it to work on a mass scale, everyone would need to replace all of their appliances, which will take at least a decade for the tipping point to occur.

Sure, but the same can be said for almost all change. Driverless cars, solar panels(renewable energy in general actually), even windows updates! I don't think saying "we shouldn't do it because it will take too long" should be a valid excuse

I hereby would like to record my claim to the term "high frequency dish washing", or HFDW as it is to be known :D

If every smart dishwasher does the same thing, the price will surge across a time period shorter than the minimum run time of your dishwasher. That stratagy only works if you're the only one doing it.

I think you are simplifying the model a bit too far. Peak load is very likely not driven by dishwashers...

It would work during the times it wasn't busy participating in a distributed cyberattack along with the rest of the internet of shit.

Hi, we are an energy company working on load shifting of EV charging. The customer will just set some preferences- the charging timing is then fully automated to shift load and smooth demand peaks. Pretty sure most modern demand response systems work this way.

How's your product going to far, and are you hiring? :P

I think you might be on to something, but this needs to be presented to the consumer in a certain way. And I do not know how far you can get without support from the big players in the appliance market.

EV and heating/cooling. Sector coupling and electrification will bring in the needed flexibility. It's not really there now with all the washing machine and appliances talk.

Cool to see you are working on this so concretely!

What is the cost of time-shifting usage versus paying peak pricing versus buying some storage to charge at lower rates? At the industrial scale this becomes a game of capital expenditures versus cash flow. At the consumer scale it's a game of convenience and cost. If energy can be made cheaper at the cost of informational overhead (or less convenience), then what is the role of storage (as it itself becomes cheaper) in managing that overhead? It doesn't seem far fetched that it will play a role.

Also, at some point, can I get a discounted rate on my hyperloop ticket if I sell it energy during peak demand at an under-market rate?

We already have internet-connected thermostats, for example. If there's enough of a price differential, the market will provide ways to do it.

  Change the price of electricity from moment to moment

This already happens. It is called the "Spot" price of electricity.

It's not just supply though, the demand, ability of the infrastructure to carry the power, the relative costs of the power plants, etc are also factored into the price.

I know what the spot price is. My electric bill, however, has a fixed price 24/7. When the cost to deliver electricity varies significantly on an hourly basis, this is not an efficient way to match supply with demand.

Note how gasoline prices vary on a daily basis. This matches supply with demand.

This is a topic I find really interesting, so I caught myself writing a whole essay about it. But I have to get back to work.

So let me just say that yes, you are right that gasoline prices better match price and demand.

The large electricity consumers, though, all are on spot prices; and in many countries you also have some split for residental users (e.g. two prices, peak/offpeak, for particular hours in the day) - adoption for this is mostly slowed down by the need to replace all the electricity meters with "smart meters" that can do the variable accounting instead of simply tracking a running total consumption.

Market-based solution? Incumbent interests object to this. Typical objections include 1) it will increase price volatility (which will hurt a) existing power generators or b) consumers), 2) it will reduce reliability. So, instead of market-based solutions, we get "energy policy" that subsidizes all incumbents and reduces innovation.

Power grids are more complicated than you think.

Markets have their role to play but it's simply an enormously complex system quite unlike anything else.

Imagine if you had to run the street system but you couldn't ever have traffic jams.

Reply on news.ycombinator.com