22 May
2019
22 May
'19
8:23 a.m.
On Wed, May 22, 2019 at 3:50 AM Christian Corti via cctalk <
cctalk at classiccmp.org> wrote:
For what we do, It's not, we don't do that, and our equipment suppliers
don't suggest it. The air is humidity controlled before it gets near the
IT racks, by building air handlers and CRAC units.
Our process water loop inside the room watches the humidity and adjusts the
water temperature appropriately. It'd be hard to for the dewpoint to
change rapidly enough to cause condensation vs how quickly systems can
increase the water temperature to compensate.
Modern systems using direct-to-system water cooling also tend to use "warm
water" cooling, since they don't need water below ~90F (~32C) to cool
effectively (some specify up to ~104F (~40C). RIP the data center that has
a dewpoint that high, since the servers will probably be covered in rust
pretty quickly.
Pat
On Tue, 21 May 2019, Patrick Finnegan wrote:
I guess I should have added "etc" to my exception list? My point was that
air handlers, CRACs, etc, should not have their plumbing running through
the data center since they should be outside in a mechanical room anyways.
Plumbing (unless you're doing aisle
containment or RDHx) shouldn't run
through the IT space in the data center.
So how exactly do you attach a modern water cooled rack system to your
cooling water system if not using plumbing?
Cooling water
to racks should be dewpoint adjusted, so you don't need
condensate drains inside the DC.
That is a requirement, you must have condensate drains just in case that
the room temperature and/or air humidity rises, or cooling water
temperature goes below the dew point.