Don't Go Being Politically Insane You Climate Change Skeptics

[sunday]

7 hours ago, Ssnake said:

On the upside, if we all - plus China and India - stop burning coal and oil by the end of this month, we still may have a chance. I remain cautiously optimistic that everybody will recognize the error of their ways. We still have a week to think of something, we can't let Greta down!

One should look at the Netherlands, and see what could be done with mitigation, especially when there is wealth in the economy of a country.

To curb, or kill, economic growth in order to keep temperature increase to 0.5 Celsius instead of 1.5 Celsius could be more devastating to Humanity than to cope with some global warming.

This approach also has the advantage of being feasible regardless of CO2 driving the temperature increase, or the inverse, and that is something that is far from being scientifically settled.

Edited June 22, 2024 by sunday

[Ivanhoe]

7 hours ago, Ssnake said:

On the upside, if we all - plus China and India - stop burning coal and oil by the end of this month, we still may have a chance. I remain cautiously optimistic that everybody will recognize the error of their ways. We still have a week to think of something, we can't let Greta down!

Heretic! Infidel!

If China stops burning coal, we will lose access to all those cheap consumer goods.

And no iPhones? Given the addiction Gens X thru Z have, that future will make Road Warrior look like Mr. Rogers Neighborhood. Imagine The Purge, only 364 days a year.

[rmgill]

They can’t handle hardship. How are they going to handle no power? 

[Ssnake]

Apple will develop a solar-powered production line for iSharePhones. You won't own your phone, but love sharing it with your neighbors in the apartment complex. Much more efficient, and better for Mother Gaia.

[Ivanhoe]

https://brownstone.org/articles/bbc-climate-disinformation-reporter-attacks-kenyan-farmer/

Quote

On 15th June 2024, the BBC Climate disinformation reporter Marco Silva published a hit piece on the Kenyan farmer Jusper Machogu, entitled “How a Kenyan farmer became a champion of climate change denial.” The reporter claims that Mr Machogu, a 29-year-old farmer with many thousands of followers on X for his campaign “Fossil Fuels for Africa,” holds dangerous views denying climate change. 

 

[Ivanhoe]

1 minute ago, Ssnake said:

Apple will develop a solar-powered production line for iSharePhones. You won't own your phone, but love sharing it with your neighbors in the apartment complex. Much more efficient, and better for Mother Gaia.

Perhaps steam generators fueled by hemp.

 

[Ivanhoe]

https://bitcoinmagazine.com/business/marathon-uses-bitcoin-mining-to-heat-town-of-11000-in-finland

Quote

Today, Marathon Digital Holdings, Inc. (NASDAQ: MARA), a leader in Bitcoin mining, has launched an innovative pilot project to recycle heat generated from Bitcoin mining to warm a community in Finland. The 2-megawatt data center, energized at the end of May, is located in the Satakunta region, home to 11,000 residents.

 

Quote

The initiative marks Marathon's first district heating project and its debut in Europe. District heating involves centrally heating water and distributing it through underground pipes to heat local buildings. Finland, known for its clean energy mix, relies heavily on biomass for district heating. By utilizing heat from digital asset data centers, the project aims to reduce carbon emissions and operating costs.

The article claims the water heated by the datacenter exhaust air is brought up to 80C then piped to buildings. If the soil temp many feet underground is about 20C, seems like there would be a lot of heat loss, requiring insulated pipes.

 

[rmgill]

That's a lot of complexity to get the water heat from the processors and exchanged with a larger cooling loop out of the data center and then down the street. Underground you'd effectively have some degree of insulation by default. 

We already insulate the water going into and out of data centers to protect everything for standard chilled water. Usually an inch or two of fiberglass and or rubber with a vapor barrier cladding. Not too hard to do so for hot pipes vs cold pipes. 

[Ivanhoe]

The article indicates the datacenter is aircooled, the exhaust air goes thru an air/water exchanger, then the water is heated up to 80C by a heat pump, before being pumped down the road.

A bit strange, but there is a chance it may work out. As with all things energy, there is thermodynamic efficiency, then there is regulatory efficiency.

I've been thinking that datacenters in northern climes could simply use their waste heat to warm up greenhouses. With watercooled processors, the waste water could be used to warm the soil in a greenhouse. Very low tech, comparatively.

 

[Tim Sielbeck]

 

[Soren Ras]

On 6/22/2024 at 11:33 PM, rmgill said:

That's a lot of complexity to get the water heat from the processors and exchanged with a larger cooling loop out of the data center and then down the street. Underground you'd effectively have some degree of insulation by default. 

We already insulate the water going into and out of data centers to protect everything for standard chilled water. Usually an inch or two of fiberglass and or rubber with a vapor barrier cladding. Not too hard to do so for hot pipes vs cold pipes. 

Using waste heat from power plants (whose primary output is electricity), data centres or other industrial manufacturing for district heating is quite well established and hardly unusual in the Nordics. The only novel bit about this Finnish exercise I can see is that it is heat from a Bitcoin farm as opposed to a regular data center.  Water pipes are insulated pretty much by default in any case, but you do need to build (or rebuild) the data center specifically for this to work and as I understand it, you need a certain critical mass in terms of DC size before it makes sense (and the bigger the better). 

 

Soren

[sunday]

The use of a heat pump to raise the temperature of the heat generated by the farm looks like a good idea. Current state-of-the-art requires a temperature so high to be unfeasible for a data center cooling system. See this chart from the wikipedia article. I guess the Finnish setup is what is called there a 3G system.

[Ivanhoe]

I know datacenter engineers have partially succeeded in convincing customers that the ambient temp of a datacenter room can be above 60F without servers melting down.

The CPU crowd has been working on server chip families with lower TDP/GFLOPs ratio. I haven't been tracking that stuff lately, dunno how successful they've been.

 

[sunday]

1 minute ago, Ivanhoe said:

I know datacenter engineers have partially succeeded in convincing customers that the ambient temp of a datacenter room can be above 60F without servers melting down.

The CPU crowd has been working on server chip families with lower TDP/GFLOPs ratio. I haven't been tracking that stuff lately, dunno how successful they've been.

 

60 F equals 15 C. Bit chilly.

If you meant 60 C, then that would be more on the line I have heard of, but still, those 60 C may be not enough for a disctrict heating application.

[Ivanhoe]

4 minutes ago, sunday said:

60 F equals 15 C. Bit chilly.

If you meant 60 C, then that would be more on the line I have heard of, but still, those 60 C may be not enough for a disctrict heating application.

Most datacenters used to have about 50F ambient temp. Very common for technicians to keep a parka at the office for times when they needed to go into the room to perform maintenance.

60C would kill people (140F) who worked in the server room for any length of time.

 

[sunday]

2 minutes ago, Ivanhoe said:

Most datacenters used to have about 50F ambient temp. Very common for technicians to keep a parka at the office for times when they needed to go into the room to perform maintenance.

60C would kill people (140F) who worked in the server room for any length of time.

 

Ok, I understand now. @rmgill could offer his expert input, but I think that is for air-cooled hardware, and only in the cold lanes. Hot lanes would be another thing.

In the case of more modern, liquid-cooled hardware, it is likely they could tolerate higher temperatures in the refrigerant.

[rmgill]

5 hours ago, sunday said:

60 F equals 15 C. Bit chilly.

If you meant 60 C, then that would be more on the line I have heard of, but still, those 60 C may be not enough for a disctrict heating application.

 

As the resident data center engineer, the temps I have always run was a 70°F ambient temperature with a supply air temperature in the 60°F range. Ejection temperatures for hardware was anything from 80°F to 110°F depending on the hardware set. 

Keeping a higher temperature based on the hardware allowing for a higher ambient temperature presumes that you have a. very redundant system and your failures of the necessary cooling don't have any substantial impact on your ambient supply air temperature. When you have a marginal reliability and you have delays in your facilities folks responding to air conditioning plant casualties, you can see temperatures rise quickly. A lower ambient temperature gives you more time to compensate for the outage and get the wagons circled to deal with the issue. 

If you have a large many CRAH/CRAC set (20 and not 3) and a very redundant HVAC chilled water plant with multiple pumps and a fast response time, it's easier to run a higher supply air temperature. 

[sunday]

2 minutes ago, rmgill said:

If you have a large many CRAH/CRAC set (20 and not 3) and a very redundant HVAC chilled water plant with multiple pumps and a fast response time, it's easier to run a higher supply air temperature. 

If a CRAC or a chilled water plant uses a water/water refrigeration cycle with a water cooled condenser, then perhaps it could be a good candidate to provide hot water for district heating.

[rmgill]

5 hours ago, sunday said:

Ok, I understand now. @rmgill could offer his expert input, but I think that is for air-cooled hardware, and only in the cold lanes. Hot lanes would be another thing.

Hot aisles can be hotter. You get the best heat transfer with a large ∆T.  

Hot aisle that are properly contained (or your cold aisles) is good for high efficiency and lower PUE (power utilization efficiency, ie the power used to cool the facility compared to the power used to run the IT load). I managed to have the best data center from a PUE ratio when AOL was checking out our data centers after we were bought/merged in the early 2000s. I had implemented a great degree of cold aisle containment which helped immensely. 

5 hours ago, sunday said:

In the case of more modern, liquid-cooled hardware, it is likely they could tolerate higher temperatures in the refrigerant.

Liquid cooled brings a bunch of extra complexities. More space demands in the racks, you get higher density, but your racks have to be bigger and deeper to account for the space needed for the water distribution and return blocks. 

I started looking into liquid cooling back in the 2010s. The gear basically drops a heat exchanger in the data center with a small, short loop so you're not running the entire building chilled water system through your computer room. Also, you want distilled water and not the building chilled water as that has larger chances for particulates. Any sort of hot taps in the building would positively screw up your heat exchanger blocks in your computers if you used that water. And if you had a leak in a rack you'd have water for days until someone found it. 

A gas cycle heat exchanger to get higher data center 'supply' temperatures is possible. In a decent air cooled data center, I guess you can get upwards of 100°F heat into your chilled water probably closer to 70-80°F chilled water return back to your chillers which are short loop refrigerant heat exchangers. 

Exsentially you have, in an air cooled data center. 

Electronics - heat sink radiating/convecting heat into the air - air circulates in the data center to the CRAH (water cooled heat exchanger) that is fed chilled water from and returns to a Chiller (Electric or gas fired usually) that is like a very LARGE air conditioner (vapor compression cycle) with both coil sets in water, the hot side is then run to building cooling tower water that is an open circuit system. 

In some of our data centers we have a giant, insulated water tank out back that has chilled water from the chiller's cold side that we can bank durning the night when power is cheaper AND when cooling towers work more efficiently. We don't have to run the chillers during the day. At Turner we ran the gas fired chillers in high summer to compensate and didn't have the giant cold tank like my current employer does. 

 

For the purpose of the Nordic system, on the chillers, aka the vapor compression cycle units, you can just set your working temperatures to be higher as you're using the water as a higher heat level carrier to go to your loads. You will have to have some way to balance when your heating demand is lower because your customers don't need or want as much hot water and your system is then giving you back still hot water because their thermostats are off. So you'd go to a conventional cooling tower or something. 

The gain is you get to sell your heat to folks who want it. It comes at the added cost of the extra-piping and the extra heat exchanger configurations. They probably have some chillers setup for feeding the hot water customers and other set up for the cooling towers or however the cool their chillers when not heating homes/businesses.  

[rmgill]

8 minutes ago, sunday said:

If a CRAC or a chilled water plant uses a water/water refrigeration cycle with a water cooled condenser, then perhaps it could be a good candidate to provide hot water for district heating.

The nit pickers (whom I work for one of) will note a CRAC is an air conditioner using gas vapor compression cycle to use freon or some other refrigerant as a heat exchange. Basically a large distributed air conditioner. 

A CRAH uses water as a working cooling fluid and doesn't effect a vapor change. 

CRAC - Computer Room Air Conditioner
CARH - Computer Room Air Handler. 

I've called CRAH's CRAC units for years or often Lieberts like people call copiers Xerox Machines. 

With either type, you're still going to be dumping the heat into something else. A typical CRAC data center will dump it into air. IF they're selling their waste heat, then they dump it into a set of water based heat exchangers. 

If you have CRAHs, then you dump it into chiller that does the same thing but keeps all your refrigerant in a confined space that is in the back of house, facilities area. 

Personally, I like CRAH system setups better. You don't end up with nasty chemicals in the data center in the event of a failure of a CRAC system failure. Be it a freon alternative and it's associated oil or ammonia. Water pipes are pretty simple and since you're not dealing with a phase change you don't have as many leak issues with a liquid as you would with a hot gas system. 

[rmgill]

Oh another consideration for data center operating temperatures is staff. Putting staff working long hours in a 110 degree data center is NOT conducive to good working environments. Those human factors are serious considerations. 

Also the heating in such cases will also make labels, adhesives and other things degrade quicker. Cables too in some cases. So you need to adjust there as well. 

[sunday]

6 minutes ago, rmgill said:

The nit pickers (whom I work for one of) will note a CRAC is an air conditioner using gas vapor compression cycle to use freon or some other refrigerant as a heat exchange. Basically a large distributed air conditioner. 

A CRAH uses water as a working cooling fluid and doesn't effect a vapor change. 

CRAC - Computer Room Air Conditioner
CARH - Computer Room Air Handler. 

I've called CRAH's CRAC units for years or often Lieberts like people call copiers Xerox Machines. 

With either type, you're still going to be dumping the heat into something else. A typical CRAC data center will dump it into air. IF they're selling their waste heat, then they dump it into a set of water based heat exchangers. 

If you have CRAHs, then you dump it into chiller that does the same thing but keeps all your refrigerant in a confined space that is in the back of house, facilities area. 

Personally, I like CRAH system setups better. You don't end up with nasty chemicals in the data center in the event of a failure of a CRAC system failure. Be it a freon alternative and it's associated oil or ammonia. Water pipes are pretty simple and since you're not dealing with a phase change you don't have as many leak issues with a liquid as you would with a hot gas system. 

Ok, ok.

So both ultimately dump the heat to air, then?

[Ssnake]

Even hot bathing water will, ultimately, shed its heat into the atmosphere, after some travel through the sewers. From there, it gets dumped via thermal radiation, into deep space.

[sunday]

8 minutes ago, Ssnake said:

Even hot bathing water will, ultimately, shed its heat into the atmosphere, after some travel through the sewers. From there, it gets dumped via thermal radiation, into deep space.

Not useful for district heating, perhaps.

[Ivanhoe]

33 minutes ago, Ssnake said:

Even hot bathing water will, ultimately, shed its heat into Texas, after some travel through the sewers. From there, it gets dumped via thermal radiation, into deep space.

FIFY