Introducing Amazon Wind Farm Texas | Hacker News Reader

Introducing Amazon Wind Farm Texas

amazon.com

144 pointsbased2110 comments

From the press release:

    Amazon today announced “Amazon Wind Farm Texas,” a new
    253-megawatt (MW) wind farm in Scurry County, Texas,
    that will generate 1,000,000 megawatt hours (MWh) of 
    wind energy annually

Seems optimistic. Rule of thumb is that onshore wind generates 30% of its plate rating. That's 660GHw, not 1000, for a 253MW plant. Still a nice, impressive number, just seems over-cooked.

It perhaps Texas better than average? 30% seems too round to really be across all areas of the country.

Yeah, that's not altogether unlikely. Kinda hoping someone more locally-versed will chime in.

Not sure of the exact figures, but whether or not the wind farm is in a constrained area or not generally determines if they can generate full out or if they will be capped at a lower level to keep from overloading the transmission grid.

We are seeing capacity factors well over 45% in Panhandle now. ERCOT rated the various windy areas of the state back in 2006. Scurry Co. was 10th or 14th if I remember correctly. Not as high cap fac as Amarillo and other areas north of Lubbock. But CREZ Panhandle is already constrained due to the nature of putting lines in the middle of nowhere with no load and no inertia. Scurry Co is on the other side of the constraint, so good move on Amazon's part. Great location with respect to wind resource and transmission access and deliverability.

For 2015 the average across the country was 32.5% for wind. [1] Texas appears to be several percentage points above the average, and newer plants often are better than old plants (if they have similar sites).

Here is state data from 2011-2012 [2], with a handy map [3]. It looks like Scurry County might be in a good part of Texas. In principle, the data for 2015 is available here [4], although one would have to process it.

[1] https://www.eia.gov/electricity/monthly/epm_table_grapher.cf...

[2] http://www.windaction.org/posts/37255-u-s-average-annual-cap...

[3] http://s3.amazonaws.com/windaction/attachments/1962/US-WINDm...

[4] http://www.eia.gov/electricity/data/eia923/

Some areas of TX have a capacity factor of 50% !

Underrated comment. Texas can't build transmission lines fast enough to export their wind power, and some utilities are giving power away for free at night because of it.

Living in Texas, while almost every provider has a free nights plan now most won't give you very aggressive daytime rates that make them worth it compared to other averaging plans (Unless you have on premis solar). Air-conditioning during the day still drives the bulk of residential consumption. I'm waiting on energy storage like zinc-bromide to mature and drop in price or something better to come along.

There is a datacenter in AZ that takes advantage of low cost night power. They have giant underground tanks full of a proprietary solution that they cool down at night, and then during the day the use it to keep the DC cool. My guess is it's just antifreeze and water, mineral oil, or something similarly boring.

It turns out this is pretty cost effective.

It should be almost trivial for a refrigeration mechanic to build a home scale system like what you have described to take advantage of free / cheap night time electricity.

It's just a heat pump with the cooling side heat exchanger immersed in glycol & water,storwge tank and a pump. Maybe some insulation. Add a heat a set of heat exchanges fir a secondary loop in to the house.

I know absolutely nothing of cooling and refrigeration, but it is an area I've been thinking of transitioning my career towards (burned out SysAdmin) and learning as a trade.

What you just described is something I actually overheard a couple of guys talking about at a neighborhood pub. A/C and Energy efficiency are bit topics in Central Texas as other posts mentioned.

Do you have any articles you'd recommend for someone with a passive interest in renewables that may cover residential projects like this?

One of my previous jobs was infrastructure maintenance in a data centre, I've worked closely with refrigeration mechanics quite a bit.

I'm at work at the moment, let me get back to you when I'm home.

Cheers!

What about the controller? Also, why not use ammonia? I can imagine that condos could also use this in places like SF.

Industrial refrigeration components are very modular, controllers capable of handling this sort of set up are available off the shelf.

The whol set up is pretty much identical to a chilled water data centre cooling set up but with a smaller chiller and a larger chilled water reservoir.

Ammonia would probably be better, I was just thinking on terms of something a fregierstion mechanic could hack together as a proof of concept in a couple of weekends.

New (taller) turbines do significantly better than 30% capacity factors. The weighted average for 2014 vintage projects was above 40% with individual projects exceeding 50%.

https://dqbasmyouzti2.cloudfront.net/content/images/articles...

We've made a ton of progress in siting and engineering of wind turbines.

Is there any energy market in the US where you can offer/buy "positive/negative operating reserve"?

Probably more related to AWS:

I could imagine Amazon getting in the energy market and trading negative operating reserve. The scenario would be:

- offer cheaper AWS compute instances on the AWS spot market

- sell the excess energy consumption as negative operating reserve on the energy market

What is negative operating reserve?

It is not easy to store energy.

So in a power grid, producers (powerplants, windfarm...) offer how much energy they can put into the grid for a certain time of day, whereas consumers, e.g. your energy provider, which manages many households, or the industry go to the energy market and say: "I need this much energy at this time of day".

So now you have to make sure this supply demand evens out.

Say for example there is more demand than actual supply, then positive operating reserve has to be put into the power grid (e.g. turn up your gas power plant). For these smaller fluctuations of supply demand you have to have something which scales quickly. A nuclear power plant is probably not so easy to scale and the output is rather constant.

Now the other case: There is not enough demand and too much energy is produced in the power grid. The energy needs to go somewhere. So now you have to find a consumer, who can quickly scale up it's demand. This might be a flour mill or here it would be AWS. And for offering that you can consume an energy X in the next hour, you get paid by the Department of Energy, whose incentive is to keep the grid stable.

So AWS would monitor the energy market and would have to know, how fast their customers would turn up more Spot instances, once AWS offers them a lower price. Then they would know how fast they can scale their energy demand in the data center and make an offer on the energy market that they can consume an extra X MWh in the next 30 minutes.

So fast scaling up demand for consuming energy from the power grid would be "negative operating reserve".

If you are just a consumer and have a consistent demand, this would not be "negative operating reserve"

Awesome, thanks

Are these wind farms located near data centers and such use the power directly? Or do they dump into the grid and sell it as power then claim all their power usage is renewable because it is "offset"?

Why would that be a problem?

Is there a flaw in treating electricity as fungible in this case? What matters is the amount of carbon that would be produced without Amazon, vs. the amount produced including Amazon.

It impacts how interesting the article is.

If they were together there could be a tech angle - like how this was going to insulate their data centres from problems with the grid, or how they were going to shed load when the wind wasn't blowing, or something like that.

Otherwise it sounds like AWS and Amazon Wind Power could basically be two unrelated companies.

perhaps AWS as a whole has a reasonable constant electric load. The price per MWh fluctuates during "on-peak" and "off-peak" hours. The price really goes up when you accidentally need more power during peak demand.

I'm wondering if wind power isn't an effort to decrease the price they pay on the open market for peak power, or to at least recoup some of those costs.

The press release itself reads "Amazon will purchase about 90% of the power generated by the wind farm." so I assume the primary purpose is their data centers.

Thanks!

They might be buying the output, but it is unlikely that the actual MW output will be what is feeding their data center. A lot of people have trouble understanding that.

Interesting because there is no Amazon DC anywhere near that area.

Yet. I wonder if there are plans to open an AWS DC in Texas.

I notice that you're simply asking a question. In your comment, you don't imply that one option is better than the other. And yet the two replies (so far) don't attempt to answer the question, but do attempt to argue with you...

Anyway, if anyone knows the answer, I'd like to know as well. And to be clear: I have no opinion about how things should be in this case, I just want to know how they are.

It was quite implied by the use of the word claim and the scare quotes.

What's the difference?

The article mentions that the previous four are grid connected but it doesn't explicitly say the same for the fifth: "The four previously announced wind and solar farms deliver energy to the electrical grids supplying both current and future Amazon Web Services (AWS) Cloud data centers."

No, this is out in west Texas where wind is plentiful along with cheap land.

Nearly all wind and solar farms inject energy into the grid just like any other generator. In an open energy market like ERCOT (where this wind farm is located) there's a market where energy is traded.

If they inject "X" units of energy into the grid and pull out the same "X" units all is good. If not (the usual case since generation and demand need to equal 100% at all times) there are rules to govern the buying and selling of energy all generators ("market participants") follow.

Suffice it to say, grid management in an open energy market is complicated. If you really want to know more about the rules you can read more here: http://ercot.com/mktrules

I would like to take a moment to appreciate that we seemingly have a pretty open, fair, and working regulatory system to make this possible. I am sure it's not perfect, but that you can do this at all is pretty cool.

One could argue that the possibility of having such structures in place is exactly what allows such a market to exist. In fact, the fair enforcement of laws and the viability of the laws themselves is perhaps one of the strongest foundations of progress in the developed nations.

yep and being a generator will probably give them direct access to this market. I'm wondering if turning some of the AWS computing power toward playing the energy market isn't some of the motivation here.

are any of the Amazon Data Centers located in ERCOT? Not that I could find in a quick search?

But grids are local in nature correct? If they inject X unit of energy into the local West Texas grid and they pull X units from the grid in West Virginia where they have a datacenter, are all things equal? I'm having trouble wrapping my head around this.

A lot factors into it, but yes -- essentially grids are sort of local.

The US has three grids: East, West, and Texas.

Folks ask why Texas has its own grid and we generally say "because Texas", but the real reason is just the way the grid grew in the early 20th century. Texas was isolated from the big industrial centers of the coasts and had abundant internal energy sources (coal and oil) so it just grew into its own grid and stayed that way.

Inside each grid the frequency is internally regulated. Each grid is out of sync with each other so power transfer between grids is limited. A tie between grids requires converting energy to DC and back into AC synchronized to the receiving grid. ERCOT has 5 of these ties and their capacity is a small percentage of the demand. Additionally, there's a few generators at the border of the grid than can operate on two different grids (one grid at a time). They choose where to operate depending on pricing.

Another problem to deal with is transmission capacity. If you have too much generation away from demand transmission gets congested. To deal with this we have nodal markets (http://www.ercot.com/content/wcm/training_courses/14/nodal10...).

> Additionally, there's a few generators at the border of the grid than can operate on two different grids

Note that this wind farm is located right on the border between ERCOT and SPP (Eastern grid member). Depending on transmission capacity buildout I could imagine them being switchable.

In 20 years I'm looking forward to buying Amazon Citizenship.

Wouldn't it be more of a membership?

yearly reserved citizenship cost .... or spot citizenship (They can kick you out of the country any time when someone is willing to pay more taxes)

> yearly reserved citizenship cost

a.k.a Prime membership.

Yes, until they send you off to Klendathu for duty...

More like "Sign-up to be a cog in our Space Mechanical Turk."

Now developers can fight bugs literally.

Comes free with Prime.

* Prime members only

The wind powered cloud.

Amazon Citizenship:

1) They setup an offshore corporation for you in a low tax jurisdiction and give you payment processors to take checks and credit cards over the internet. Every few years, they move all of these corps en masse to lower tax jurisdictions, if any. More efficient competition among tax regimes. Amazon will file all pertinent taxes and paperwork and connects to your quickbooks to make that even easier.

2) They provide a phone number to an Amazon Embassy hotline. If you are ever captured by anyone and local authorities cannot help, call this number and Amazon PR will lobby for your rapid release and return to safety through aggressive social media outreach. They will also setup a gofundme page for you to cover legal expenses, etc.

3) Give up your American citizenship (if any) so you no longer have to pay taxes in the US while not living or working in the US.

4) Amazon World Citizen Passport

So you can work, maybe get some physical/political help if that's needed, and travel freely. What else is there?

The cost? 5% of your income.

Traditional governments would ideally end up selling physical living permits, allowing you to live physically in a given country/area/union, keeping the rest of their tax laws intact for anyone still incorporating or working within the given country. IE, no existing laws really need to change. The price of these permits ends up being a function of location desirability, access to services, and relative guarantees of physical safety. The job of governments becomes to improve each of these variables.

Bring on the Snow Crash-esque future.

I'm surprised we don't have that happening in places like China or Saudi Arabia. I wonder what the Huawei compound would look like.

It's already here.

Just look @ Snapchat HQ. There are homeless people sleeping literally right outside the main entrance. The epitome of high-tech, low life.

Given what it's like working in their warehouses, this will be a two-tier civilisation, like many that exist today.

Dear Citizen,

Due to too many searches and queries on your Citizen account recently your account has been closed with immediate effect. Your Citzenship is now revoked and your Amazon World Passport is no longer functional. The Embassy Help Request you opened five days ago has been closed.

We're sorry for any difficulty caused, but we strive to keep costs low for all Amazon World Citizens.

Please note this decision is final.

Amazon.

PS All of your "digital property" is now gone. Goodbye.

To be fair, that can happen already given current citizenship schemes. And if the key service provided by a "citizenship provider" is filing paperwork and running PR for groups of interests, there's no reason to assume a monopoly. If governments competed to make their citizens lives materially better, that might be better than what we've got.

On a somewhat unrelated tangent, there's that famous kennedy quote, "Ask not what your country can do for you, ask what you can do for your country." It's often cited as being this profound political statement. But it's not. It's basically the definition of an abusive relationship. A lot of people believe citizenship should not (or cannot) be a fluid concept directed by competition and self interest. I generally disagree. But I'm not talking about libertarianism; I don't think government needs to be abolished. It's more like, they need tangible goals and metrics and accountability.

I'm sold. Where do I signup/send my money?

You print your passport yourself, because Bezos is proud of the cheap culture he creates.

I'd swap cheap with efficient.

Clearly the Presidential title would be: First Prime

I doubt this will offset the huge boost in emissions which will come with Amazon's 'Flex' service, which will see products delivered fast via horribly inefficient passenger cars, as opposed to relatively efficient vans and trucks.

One person delivering to many people will be more efficient than many people making the trip to the store in their own cars.

Yes that is true, but why go backwards?

Because the current van and truck experience for me is a tracking update at 6am saying "Your package is out for delivery" and then it actually gets delivered sometime in the next 12 hours. As a customer that's a crappy experience. Hopefully flex will push everyone to up their game with regards to delivery experience. If vans and trucks are more efficient then they should win out in the long run because they can operate at lower margins.

Disclaimer: I work with Amazon, not on the Flex team. My words don't represent the company.

> sometime in the next 12 hours [...] a crappy experience

How perspective changes. I remember when ordering anything meant "4-6 weeks for delivery".

The 12 hours here refers to the delivery window on the delivery day, not the time to ship the package. Though the total delivery time is certainly much less than 4-6 weeks.

The 4-6 weeks is a 2 week delivery window! (No tracking numbers then, either.)

I don't understand your comment. Do you mean that we should back to the 4-6 weeks delivery or if you're just annotating the notion that things have improved in the the last few decades?

It's that a 12 hour window being "crappy" is like saying your transatlantic flight being delayed 30 minutes is "crappy" (my first Atlantic crossing took 10 days I think). It's all about perspective :-)

Once we have a taste of what's possible it's really hard to accept that the status quo is good enough. With Prime Now I can get a lot of things ordered, shipped, and delivered within 2 hours, with live updates when the delivery person is on their way and arrives outside my house. I remember the days of multi-week ground shipping, I'm really glad we kept pushing the limits.

It still does - I'm in Australia.

Will this wind facility actually be connected to an Amazon facility, or is Amazon doing this to collect the wind production tax credit?

Oilman T. Boone Pickens almost made a big investment in Texas wind in 2008. Could be he missed the boat.

http://www.cbsnews.com/news/pickens-plan-for-huge-wind-farm-...

Pickens Plan actually did invest in huge wind farms. Not sure about in Texas, but Oklahoma is littered with his windmills.

Don't forget part of Pickens' Plan included water rights from West Texas to Dallas under those wind farms. If I remember correctly, there were some disagreements and TBP wasn't able to complete the water pipeline portion.

He didn't miss the boat. His master plan of connecting West Texas water to Dallas for marked up resale fell through.

https://gigaom.com/2008/06/20/t-boone-pickens-taps-water-win...

T Boone was too far ahead of the transmission infrastructure with his plan. The SPP lines were not capable of moving much wind (pre ITP build-out) and the ERCOT CREZ plans were just a twinkle in our eyes at the time. We came up with ERCOT CREZ prelim plans about 2006 but it took until 2013/14 to get em built and energized. So, he launched his boat too early, he did not miss the boat.

It seems to me that the long term value of wind farms is locations with a lot of wind at night, to keep generating electricity when solar systems shuts down for the night. This helps solve the energy storage problem.

There will be a few hour lull after Texas solar goes dark and before night wind kicks in. In Texas, all of our wind is night peaking with the Gulf Coast having a significantly better curve for servicing ERCOT day load. So storage will play a very big role in ERCOT as we move more towards renewables.

Can't wait for Amazon Energy... Utility-scale renewable energy for all

No doubt bundled with Amazon Prime.

Stop trying to hoard the air!

Serious question - as of right now, does Amazon have wind farms to reduce their own costs or do they plan on making a profit by selling the energy to consumers?

Lincoln Clean Energy owns and operates the wind farm. Amazon is in a long-term contract to purchase 90% of the energy produced. It's open what happens to the remaining 10%. I'd say Lincoln Clean Energy sells those and any energy Amazon doesn't need to the grid.

If this is true, the article is pretty disingenuous. It reads like Amazon is building the wind farm, look how dedicated we are to renewable energy, we're investing actual money. If someone else is building it with their own money, and Amazon is simply buying the energy they produce, that's kind of a big difference.

Wouldn't it be the same difference? They either profit by offsetting costs (selling the energy) or they profit by reducing costs (using the energy). Energy is fungible, right?

I'm going to say never. The energy demands of data centers will be skyrocketing due to increased demand for machine-learning, those GPUs and ultra-multicores will be hungry for decades to come.

It could just be a hedge against future taxes on fossil fuels which seem inevitable if we are going to meet the Paris commitments on climate change.

electricity is a very liquid market, they're probably selling this generation into ercot, buying it in offset in whatever grid they operate on. they could have some sort of physical exchange in place to act as the offset (i.e. they sell power to a generation company in Texas, they buy same offsetting amount from same generation entity in California/MISO/wherever)

Honest Question: Is this a profitable endeavour? or it's just for the positive PR?

I wonder why they went for Wind farms over anything on the sea or solar powered

They mention their existing solar capacity in the article. Presumably this will also grow over time.

Doing anything at sea would surely be 10x as complicated and expensive...Texas has wind and open space and roads.

Solar is more suited for 24 hour power, which Amazon needs. It also costs less than solar, but needs maintenance, but if you're vertical integrated like Amazon then having your own low cost maintenance team is probably not an issue.

There's probably also a bit of "Apple and Google have solar and fuel cell covered, where can we make an impact?" involved. Not a bad thing. We need to be investing and experimenting with lots of different alternative energy strategies.

I would like to see someone do a large scale implementation of marine current power (underwater turbines spun by currents like the Gulf Stream).

No big deal. I knew it was only a matter of time, since it was already announced on Parks and Rec a while back. :-)

https://www.youtube.com/watch?v=XFKoGtgg6Mo

Hopefully they can cut down on excessive packaging at the same time: packagingnews.co.uk/news/environment/amazon-customers-criticise-excessive-packaging-21-12-2015

Seems that in a single swipe Amazon would more than double the total USA’s wind generation capacity when this goes into production in 2017. A really fantastic achievement (or am I missing something in the data?).

Total US wind energy in 2015: 73,992 megawatts [1]

Amazon’s Scurry County Texas capacity: 1 million megawatts

[1] http://www.greentechmedia.com/articles/read/7-Charts-That-Sh...

Edit..

From the replies it seems it would be at best ~0.3% of total wind generation capacity in USA.

They listed MWh/y which is a pretty strange unit. The farm is 253 MW, so not a significant percentage of total US generation.

Source: http://phx.corporate-ir.net/phoenix.zhtml?c=176060&p=irol-ne...

Nothing strange at all about MWh/year. A MWh is just 1000 kWh, which is a very standard unit of energy.

With wind farms it makes more sense to talk about it's production in MWh rather than it's capacity in MW. The actual energy produced by a 253 MW array will vary significantly based on where it is sited, etc.

hours/year (= 1/8765.76) is a strange factor by which to multiply a perfectly good SI unit (MW).

Objectively it is strange, but it is the norm in that industry.

The watt (and hence kW and MW) are measurements of power, not energy. kWh (and MWh) are common, internationally accepted units for measuring energy.

Wind (and solar) energy production tends to vary seasonally, so it makes sense to use a year rather than some other time unit when discussing how much energy these things can produce.

I upvoted you for contributing to the discussion, not for being correct.

It looks like Amazon is adding 1 gigawatt across five projects, to the existing 75 gigawatts.

No, "1,000,000 megawatt hours of wind energy annuall."

/ 365 / 24 = 114 MWh per hour on average. Or ~250 MW nameplate power depending on capacity factor.

Ed: They list 253 MW.

They imply 45 cap factor with the 1,000,000 MWh which is expected for Scurry Co and the newer turbines. 1000000/(253*8760) ...rounded ignoring leap year.

That's one million megawatt hours. Watt-hours are the most confusing unit ever, and measure energy, not power. (A watt-hour is the amount of energy produced by one watt for an hour.)

1e6 MW hours = 3.6e15 joules. The power plant generates this over a year, so that 3.6e15 joules / year = 114 MW on average.

Reply on news.ycombinator.com