Well, that's the thing that determines the pipe thickness. As it's circulating water, it's 100bar in the down leg too. The pipe had better be well-lagged, otherwise it will be warm water at the top.

Maybe the installation could be similar to that in the CN Tower, in Toronto, only a bit bigger?

<lateral thinking mode>

Maybe chilled air could be circulated instead?

</lateral thinking mode>

Seriously, I wouldn't go up there without a parashoot. S.M.

Most hydraulic applications related to big height differences, is designed so that absolute pressure in ALL positions, is within economical boundaries, i.e. pumping power needed for it, is distributed in (height) stages. Every stage takes care of a limited height difference. Not as simple and foolproof as it might appear, but it works. S.M.

Big deal. They'll just use Schedule 160 pipe for the first 150 floors, 120 for the next 100, 80 for the next 100, and 40 for the rest. And a whole bunch of pressure regulators.

That's right SE. The building will have more than 1 plant room at different levels.

They're not going to have a single system for the whole building.

Booster pumps. You don't even need the cisterns, just locate your pressure gradients appropriately (which should be child's play for whoever's building this !).

It id VERY likely that the primary chiller/tower system will be at the TOP of the building (where the towers usually are) so those devices, including pumps, etc. will be sized for "normal" pressure.

Air handlers (either by floor or groups of floors) will be served by "local" pumps circulating a secondary loop fed through a high-efficiency heat exchanger on each active level, again- keeping local pressures "normal".

Another BETTER option, still with key machinery on the top floors, is to use refrigerant with the down-feed being liquid and the return (gas) going up, again with AHU's per floor or group of floors using DX coils. Totally eliminates lots of BIG, thick pipe and is a lot cheaper to install.

Assuming that they use an oil separator on top, or just HXU's to condense and cool the refrigerant, they will not need any return oil lift bends, just straight pipe. They could serve the HXU's with a small loop of CHW or even just another refrigerant loop with air-cooled condensers off the compressors- which would be the ultimately BEST choice because water is a fairly high grade commodity in that part of the world and it is the lightest weight. The slight compressor efficiency hit would be over come by the reduction in condenser and chilled water pump energy.

That's one approach.

My former boss designed the chiller system for the World Trade Center. The (7) 5000-ton R-22 chillers were in the basement. After the 9/11 attacks, there was some question about dangerous pressures still remaining in the system. IIRC, the system turned out to be "flat" (i.e. discharged via some leak).

Some of this concern was described by James Fallows in "The Atlantic [Monthly]" magazine.

(See also Lyn's "I Flew Today" thread, in memoriam to that event.)

There will be void decks at vertical intervals to accommodate the plant. Just as they do now in <1km buildings.

Clearly visible here on the 60 story MLC building in Sydney.

and the AAP building also in Sydney.

Look up next time you are in a concrete jungle and you will start seeing them everywhere.

Haven't you ever wondered why it takes longer to transit between some floors when you are in an elevator in these (or similar) buildings?

water is pumped to the first mechanical floor. appox. 100'', then an other pump pushes it to the next mechanical floor, and so on until it reaches the top floor.