Comment by belorn | Hacker News Reader

Comment by belorn | Hacker News Reader

Parent Full thread

Two major problems that need to be considered with using localized pricing are sudden bidding wars and partitioning.

Let's say locally you have a few factories when a sudden drop in supply occurs in the energy market. The biggest costs of those factories are personnel, equipment and building, so rather than shutting down they will be willing to pay much more for energy than what is rational in the market in order to continue producing and maintain contracts in the short term. Long term they may move away to a more energy stable and cheaper location, but for the next 10-20 years they will eat the occasional spike in price. Everyone else who locally live there can't eat a sudden 100x in energy costs, with the result that they will vote into office politicians that protect against such situations, and those politicians will in turn spread out the problem over a wider area where areas with stable supply of energy can help through funding (higher energy price) and transmit energy to areas with less stable energy. Alternatively they can pay their citizens' bills directly, as happened during the energy crisis a couple years ago, money that then get taken from the general budget (paid through taxes).

The other issue is partitioning. The more local zones you get, the more borders you get between those zones. People living on the border will see a unreasonable price difference depending if their house/town/city happened to land on one side or the other. This feels unfair, which results in upset citizens, resulting in people voting in politicians that can fix the situation. Politicians then feel a pressure to even out the prices among the zones, using things like adjusting local taxes.

As a general rule, the voting population want stable energy availability, stable prices and fair prices that are similar to everyone else.

Power demand is often inelastic. There are many places in summer where people would pay $200/kWh to run a small air conditioning unit, for example, when their normal price of power is $0.10/kWh. There is no reason to shift toward generation sources that have price uncertainty when demand is inelastic like this. See the exact same thing in healthcare. The proper solution to this economic problem at a societal level is to have a predictable low price of available power.

There are "many" places where people would pay US$200 to run an air conditioner for one hour?

Hospitals, specialty warehouses, datacenters, the list goes on. Some private citizens in Arizona or Texas, too.

> There are many places in summer where people would pay $200/kWh to run a small air conditioning unit, for example, when their normal price of power is $0.10/kWh.

I don’t know anyone that would pay $200/hr to run a 1kW air conditioner. I’d just go in my air conditioned car and pay for gasoline..

I once had to pay about $1500 USD per kWh. That’s $90 USD to boil 1 litre of water.

Consequently I chose to freeze rather than turn on any heating. At that price even LED lighting is too expensive. I probably should have unplugged my fridge too.

But many people chose to keep themselves (and families) warm.

Lesson learnt: never pay the spot price for power. In minutes I probably lost all the saving I had accumulated by micro managing power until that point.

I have seen my fair share of data centers and office complexes switching to their diesel generators due to very high energy spot prices.

If you have an alternative source of power (even if it is a gas generator) I think a spot price contract is fine. Otherwise it is too risky for residential consumers.

Good point. If there is an alternate power source, it makes perfect sense to go to spot pricing. If you don't and you have room mates who may not like freezing in the middle of winter, don't.

I would only be on a pricing plan like that if I had a way to automatically cut off my service if price exceeded some limit. If this made power outages common enough I'd install batteries to tide me over most of them. Functionally this would be equivalent to an unreliable power system.

Ultimately, "shutting down the house, draining the water pipes, and going elsewhere" is a solution for extreme winter events.

Parent may be referring to specialty units that for example chill expensive perishable medication.

That sort of facility will have diesel generators for backup power, they will be maintained and tested monthly, load banked yearly.

Even if the generators failed, renting a trailer mounted diesel generator costs much less than $200/kWh.

If you really want to turn everyone who runs a facility that needs power for under $10/kWh into an energy trader who has to build a facility with a larger generator than needed and a battery bank and everything else, there is a huge amount of economic loss involved in that. Energy is cheap. Energy should be cheap. That infrastructure should remain for disasters.

Running those generators is not free, and puts wear on them. Also, starting up those generators takes time and puts a certain amount of wear on them, so you are now talking about also having batteries and other equipment to cover short periods and the gap during generator startup. And now when do you spin up the generator? Do you spin it up when the instantaneous price of a kWh goes over a certain threshold or do you wait? All of these are now things that every single industrial facility has to think about.

In my post I was explicitly referring to backup generators being basically mandatory at a refrigerated pharmaceutical storage warehouse, not making a general argument that every industrial building should have backup generators. Generators are common at data centers, hospitals, emergency dispatch call centers, and refrigerated storage facilities.

There was no need to take a specific point I was making about refrigerated storage facilities and pretend I was talking about commercial and industrial sites in general. Please read more closely next time and respond to my actual argument instead of the argument you imagined that I made.

My point is that running those things is not free or cheap, and that with floating energy prices, you now have to make them significantly beefier and make economic decisions about when to run them. You will also have many more of these generators show up. People adapt their behavior to the incentives you give them, and the design of generators is also very much more subtle than you are giving credit for.

So yes, while it seems like people already have infrastructure for floating energy rates, in reality they do not.

Again, I am not saying that facilities should use backup generators to make power if it spot price goes slightly above what it costs to run the generators. Read the original post I was responding to, all of my posts in this discussion are operating under the assumption that electricity is $200/kWh.

Let’s say you’re running a 250 HP compressor to keep your cold storage warehouse cold, that would draw around 200kW. One hour of operating that would be $40,000 if electricity was $200/kWh. It costs much much less than $40,000/hr to operate a 250kW generator, you can buy a 250kW diesel one with an automatic transfer switch for around $100k from Cat or Cummins.

I’m not replying if you respond again, you’re responding to straw men instead of my actual arguments. I sell and run commercial electrical work, including generator installs and replacements, I know everything you’re telling me already.

If you can't handle volatile prices and know your demand in advance then simply by PPAs or energy futures?

You know, the standard method all markets use to handle volatility. But apparently electricity is different and we need enormous subsidies instead.

It's simply not realistic to expect consumers or most small businesses to participate in futures markets. And futures markets sometimes break down with the counterparty failing to deliver. That's not a problem with most commodities but electricity shortages cause real problems.

I expect the only viable solution left is to go around the meter using an aggregator like Tesla Autobidder. A large entity consolidates many home batteries as one "virtual" battery, handles the grid futures prediction and dispatching for a cut, and re-distributes a majority of revenue back to the battery owners.

This effectively uses the existing behind-the-meter grid market to make an end-run around current perverse (non-local, non-instantaneous) end-customer pricing schemes.

A downside to these schemes is that the stability of the electric system now also depends on large communication networks and some computer solving some algorithm. I can imagine the outages. I prefer my power outages to be caused by good old fashioned weather rather than Hadoop kubernetes cluster operator eventually consistent race error.

Yes I agree. Ideally there's an ecosystem with many providers, minimizing the impact of any one system failure.

More likely, the NSA / Homeland Security will assist large providers in system hardening and resilience.

They already do - here in the U.K. you can fix your energy price for 12-24 months typically which is a gamble in the future price of energy.

Though in that case the exact prediction made did come true. Massive gas price increases due to the Ukraine war bankrupted a bunch of businesses selling fixed price energy that they could no longer supply without making unsustainable losses.

Various government interventions followed but they were so bad you might assume that they were more interested in ensuring their mates in the gas industry made high profits than protecting the nation from the fallout.

I suspect the real failure mode was around the price cap. Utilities should have been hedging against wholesale prices raising above the cap for a sustained period of time as occured. Net-net the result is the same and the cents in the dollar saved when prices were normal / stable were long gone when prices went much higher. (Note in the UK at least we have been in payback for a long time now with customer pricing at the cap and wholesale typically much much lower)

Electricity is a bit of different in that it is an general utility, closely attached to government for delivery and production, and historically dependency on natural resources or caused pollution (social impact). Electricity is also essential in many places for human survival, and is critical for economical growth and social stability. During war, electricity is treated different from other commodities in a very clear way.

But my point is that all that isn't necessary the important part. The voting population are not always rational participants in the market, nor are companies. There are Nash equilibriums and strategies with local maximums that results in irrational consumer behavior. When there are major social consequences from irrational behavior then people will look towards social, ie political solutions. That generally means regulation, subsidies and if all else fails, government control. The only way to avoid that is to either eliminate the social consequences, or eliminate irrational behavior.

I think it is different than other commodities since it is very expensive to store electricity. A diesel powered generator and tank full of diesel is probably the best option.

Everyone adapts to the single most expensive uneven form of local pricing: property prices.

I think there's a very strong case to link grid and generation capacity improvements to local pricing in the effective area, even if it doesn't end up being through the market. Make the link explicit: "if this pylon project is built, we will give everyone in the receiving county 1p/unit off their bill for five years."

“Nodal pricing” exists in several markets but isn’t often used. One example is in Ontario the nodal price in Thunder Bay was always very low since industry there had collapsed but there were still three coal fired power plants. They could bid a low price so it would be accepted, and due to low local demand the supply and demand solved for a low price in Thunder Bay, but due to transmission constraints all of the bidders could not generate what they bid, so they would have to be constrained off. I can’t recall if they got a payment for the power they weren’t able to generate.

People only have to think about property pricing every several years, when they move. Electricity spot prices change by the minute. So the mental burden is quite different.

Reply on news.ycombinator.com