Your Air Conditioning is central to keeping the lab environment stable, the same as with other labs around the world. That’s not just a matter of convenience: experiments perform differently at different temperatures, cells propagate faster or slower with changes in the temperature, and humidity can wreck everything.
An A/C system for a lab will put air in and take air out. However, the A/C air supply isn’t the only way air gets into the lab; it will also take air in through cracks, under the door, around windows – there simply isn’t such a thing as perfect air insulation (read the story of Biosphere Two for further info). So, the lab system works like this:
Air Removed by A/C = Air supplied by A/C + Air supplied from elsewhere (“offset”)
That balance is really important. For safety’s sake, your lab should be always in negative pressure, that is, there should be more air going out than coming in. That prevents anything nasty from spreading; instead it is sucked out through the vents. However, if these two systems get too much out of sync, bad things happen.
One phenomenon I’ve seen is the “Hurricane door”. This happens when the air supply is severely decreased or turned off. A lab like ours can move through 50,000 cubic meters of air each hour; when none of that air is being supplied by the A/C system it is sucked in under the door or wherever it can. Since it is being pushed through smaller cracks, you get a howling gale; it can get so bad I need two hands and all my body strength to even open the door.
There is worse though. Air supplied to the lab is first cooled down to remove all moisture, and then warmed up again to the requisite temperature (ironically, it is more energy efficient even in hot climates to have the temperature low than high). If dry air isn’t being supplied, you can get a gale of hot, damp wind into a cooler lab. The effects happen in minutes: condensation fogs up all glass surfaces and water begins to pool on desks.
The converse situation – when there is air flowing into the lab but not being removed – is less risky, but still a problem. You can get doors refusing to shut under the pressure, and you lose your main defence against gas spreading. This is still better than having no air conditioning, or only having air flowing out, and it can be useful if you have 1) secured all gas cylinders, and 2) have cleared the lab. So it can be a useful stand-by during maintenance operations.