eigenvector

Yes, this function exists in some fashion in every grid (although the economic setup of how you're paid for that service differs widely).

The problem in ERCOT is demand massively exceeding available supply over a period of days. You need to carry your primary operating reserve all the time, because you never know when a generator or transmission line may trip.

When there's simply a lack of energy supply to meet demand on a steady state basis, you still have to preserve your primary reserves for other unexpected events and cannot call them in to meet longer term (30+ minutes) demand. So load shedding happens once available reserve dips below a certain threshold (generally around the single of the largest single contingency on the grid which may be the largest single generator or a line that feeds multiple large generators).

ERCOT hasn't run out of primary reserve, but they are going to have to shed load in order to maintain adequate reserves. You always have to shed load before exhausting reserves, because one moves you closer to stability while the other moves you further away (consuming operating reserve).

At this point, with power prices spiking to thousands of dollars per MWh, any plant that can run is running, including some that have come out of planned outage earlier to help out (and be paid handsomely). Even if you simply invoked some emergency order that "all generators must run until further notice", there simply isn't enough capacity due to the large number of generators on forced outage. There's no way out of that situation except temporary load shedding.

ERCOT is pretty good about releasing their event analysis after the fact.

I think the main issue with renewables is not inherent unreliability but that grid operators do not understand the potential contingencies as well as with conventional generation sources.

SPP has put out some good analyzes of recent max generation events there involving large amounts of wind being unavailable in cold snaps. It turns out no one at SPP bothered to ask wind turbine owners what temperatures they were capable of operating in and some wind operators didn't buy the cold weather package for turbines despite installing them in the Midwest. So generation availability forecasting didn't take into account the real capabilities of generators - a clear failure in operational planning.

This is more of a regulatory issue than a technical one. The machines can operate at -40 C, but someone has to actually spec that functionality and ensure that the associated subcomponents (heaters etc) are properly maintained.

There is a new NERC reliability standard in the drafting stage for this.

The technical term for what you're describing is under-frequency load shedding (UFLS) and it is part of most modern power systems around the world, including all of the North American grids. It is an important safeguard against system collapse, however, the goal of system operators is to avoid reaching this stage. In fact, resolving contingency events without interruption of "firm load" (that is, loads that have not previously agreed to be interrupted) is an important performance metric for any grid operator.

UFLS is mostly intended to buffer transient loss of power supply (for instance, tripping of several large generators). If the grid operator knows they are facing an inadequacy of generation supply and all resources have already been called in, they will start to shed load under manual operator action to avoid UFLS activation. Before doing this they will declare an emergency which, generally speaking, requires all generators to make best efforts to supply as much energy as they can to the system.

Think of it like automated emergency braking on a car vs driver braking. AEB is great, but if you can already see that you're gonna hit something, just hit the brakes right now instead of waiting for AEB. By the time UFLS kicks in, you're already in dire straits and have only moments before reaching an unrecoverable state.

The base capacity is 64 GB now, but lots of phones of 32 GB and even 16 are still usable (the original iPhone SE, for instance, was sold in a 16 GB version and runs the latest iOS).

These older phones are way more common outside of the US and even just outside of the SFBA where not everyone buys a new $1000+ phone every 2 years.

I agree that of all developed countries, tipping is most prevalent in the USA, but as a fellow Canadian I wouldn't assume our Amazon delivery people are being paid anymore. They are independent contractors and not subject to minimum wage laws. If anything, their effective wage might be lower because vehicle operating expenses are higher in Canada.

For a long time this sort of worked for both sides as the bathroom foot traffic would often buy some small item in the store and most items sold at gas stations have a hefty markup. Buying a Coke for $3 might net as much profit for the gas station as buying $30 of gas. Besides, the gas station can never tell if you're going to buy gas and just want to pee first.

Since the pandemic, a lot of gas stations summarily closed their bathrooms - even to customers. It sucks.

The Canadian government paid Deloitte $16M for exactly the same thing, with a deadline of January 31 to get it running.

https://www.ctvnews.ca/politics/ottawa-gives-accounting-firm...

One of the factors cited in awarding the contract was Deloitte's "experience" doing the same thing in the USA.

http://www.canadian-accountant.com/content/business/why-delo...

Now think of the problem in reverse, from the government's side. Why are they contracting it out when all of the most difficult aspects involve integrating with government systems, policies and personnel - things that government already knows how to do?

"Because we don't have the expertise to do software development."

The comments in this thread, which I fully agree with by the way, just cement my belief that if government wants to deliver high-quality IT projects they need to start developing the expertise to do it in-house.

A seaplane flight over the west coast of Vancouver Island from bases in Tofino or Gold River is a wonderful and affordable flight-seeing excursion as well (around $500/hour total for 3 passengers).

Nothing inherent to cell phone networks causes them to go down during power outages. Just like POTS they need a backup power supply for their infrastructure to remain functional when the power grid is down. Usually this takes the form of diesel generators.

The main difference is that the cell network doesn't power the handset but modern cell phones use so little power than a $50 consumer UPS would recharge a cell phone many times over.

Keeping phone networks up during power outages is very important for public safety.

Hi there, I checked out your website. It's a cool idea and one that is needed.

What do you think of the fact that many of the products are in fact disposable by backed by a "lifetime" warranty precisely because they're so highly marked up that the manufacturer can give you a new one for free instead of attempting even trivial repairs? Osprey is the king of this. This seems to be the opposite of the buy for life mentality - things should be repaired if economical, but no one repairs what they can get replaced for free.

Indeed. I've had the good fortune to pass long periods of time (from days to months) hiking and it's amazing what entertainment just meeting a new person after weeks can provide. And no matter what book you bring along, you will finish it in a couple days without constant distraction.

It's not just 5V, though. PD can supply up to 20 V and poor implementation can brick 5 V devices by supplying the wrong voltage.

The absolute worst was a Dell laptop I had around 2008 with a combination USB-A/eSATA port. If you tried to plug a USB connector in upside down, you were basically connecting the USB pins of the device to the SATA part of the port and the computer would inexplicably reboot.

If you think USB is bad, try looking at all the various things that can be connected together with M.2 slots vs. the pairs of devices that will actually work together despite having the same slot. I have a degree in electrical engineering and after years I only just realized that there are actually PCIe M.2 storage devices that only support AHCI mode not NVMe.

You're right, but it's important to note that offshore is only about 1/10 of the global wind power industry.

The "who has the biggest turbine" race between wind turbine OEMs is the Formula 1 of the wind industry. It's mostly about bragging rights and long-range R&D. The vast majority of new turbines being installed today are in the familiar 2-4 MW class. The >10 MW turbines are more important as an R&D platform than as an actual contributor to energy production.

4 MW class onshore turbines first introduced in 2017 are just now starting to hit their stride and be installed at large scale.

Well said. To paraphrase Wade Davis, the question is not hydro or no hydro but where, how much and for whose benefit. Many dams produce very little electricity and can be removed with minimal impact to the cost or carbon intensity of our power supply.

We should not conflate dams that are integral to the power system like Grand Coulee with those that are insignificant.

From a hydrotechnical perspective, it's a feature not a bug. You do not want sediment going through your turbines. Suspended sediment is a major maintenance issue for run of river hydro plants.

From a security perspective, ChromeOS bothers me because it has a much shorter software support lifecycle than the hardware lifespan.

We have ChromeOS devices out there are perfectly able to run the latest ChromeOS in terms of specs but aren't supported anymore by Google because of an arbitrary time-based EOL and therefore will either be browsing the web unpatched or thrown away.

Whereas any other x86 computer from the last 12+ years or so can have up-to-date Windows or Linux put on it.

Another step on our march to disposable everything.

In almost every developed country in the world except the United States, it would be illegal to sell a defective item then refuse to refund or replace it, whether it's $1 or $10,000. Consumer protection regimes are a real thing that exist, although maybe not in your jurisdiction.

So yes, you can legislate that. By making merchants bear the cost of defects and take a loss on replacing the product, you make it no longer a viable business model to constantly sell broken stuff. It's a proper alignment of incentives.

In a world where Google is being sued for antitrust in part because of a monopoly they built by giving stuff away for free, perhaps making it harder to have a "low margins, but we aren't accountable for anything" business is not a bad thing.

You're also building a strawman - that Google would need to spend a lot per user on customer service to start actually providing customer service instead of refusing to do so for most of its products.

It would need to spend a lot in absolute dollars, sure, but the vast majority of users would have precisely zero support interactions. No one is saying that Google should have an army of 500,000 support reps. Just that it should actually be possible to contact Google without knowing someone who works there.

No, at a basic level how much they make per user is not the user's problem. That's Google's chosen business model.

You cannot sell a defective thing then when the customer comes back say "sorry, we actually only made $0.01 on that so we don't provide customer service, goodbye."

Your obligation to not, for instance, arbitrarily revoke someone's account access causing them measurable harm is not dependent on whether you deem yourself to have a sufficient level of profit on the transaction.

Companies have obligations that external and not bounded entirely by their commercial whims.

It's also much easier to do hard things when you're making billions of dollars.

I find the argument that Google et al cannot -afford- to do anything but treat their users in an arbitrary, high-hanged way unconvincing.

Respectfully, as a power systems engineer, I'm unable to grasp how what you're describing is beneficial compared to current power networks or how it respects the underlying physics of electricity.

The design of power transmission and distribution networks is driven by several physical realities: high voltage is needed to transmit power over long distances with acceptable losses, it is easier and cheaper to transform voltage and switch ultra high power circuits in an AC network than a DC network, all power conversion is lossy, electricity is hazardous and therefore electrical circuits must be insulated and protected adequately to protect public safety.

I've heard this question of "why don't we have some kind of universal power outlet that any load regardless of voltage or current needs can connect to?" before. The answer is the same as in many other fields of engineering: it's not efficient to overbuild every node in an network to handle any possible load when electrical loads are usually well-defined and do not change rapidly. We don't have a universal train that runs in the metro, for freight and for high speed rail not because it's not possible but because that would not be efficient. The same applies here.

It only costs about $50 USD for a battery replacement from an authorized Apple service provider. If you're satisfied with the device otherwise, a new battery will keep it going for far cheaper than a new device.

You appear to be conflating Intel's 1990s antitrust woes and the success of the Core microarchitecture which launched in 2006.

Despite Intel's well-documented anti-competitive practices, AMD had decisively gained the upper hand in performance by the time of K8 (2003). Intel's microarchitecture at the time, NetBurst, was uncompetitive and relied almost entirely on inertia, and yes, anti-competitive behaviour to retain market share.

Core leap-frogged that and Intel remained ahead for the better part of a decade afterwards. To say that Intel sat on its laurels between the time of the first round of antitrust litigation in the early 90s is quite inaccurate.

"Beyond the Hundredth Meridian: John Wesley Powell and the Second Opening of the West" is a fascinating look at that topic as well.

ZIP code verification is commonly used for credit card authentication at gas pumps in the US, especially before the advent of chip & PIN.

This is immensely annoying for Canadians who drive their cars to the US, although there are some built-in workarounds supported by the credit card processor such as punching in the 3 numerical digits of a Canadian postal code followed by 00.

I completely agree that their methodology is not robust.

But Amazon also refused to state what fraction of items go to landfill, despite making subjective claims that it's "on occasion" and the "vast majority" are liquidated or resold. If that was true, Amazon would have the data to back it up and would only benefit them to say so.

If we want to know the actual number of items that go to landfill, we don't need to send GPS trackers everywhere, we can just mandate Amazon (and all other companies) to report on that. It's not like they don't have the data.

The flip side of this is Amazon actually refuses to inquire with delivery companies, and here in Canada, the post requires that the sender be the one to inquire.

I've had several instances where the item was clearly at my local post office but misplaced. Canada Post refused to check without Amazon requesting it, and Amazon refused to do anything but send a new item. In each case the original item mysteriously appeared in my box months later when the postal workers inadvertently found it.