Make sure you point out that with the AC failing (not sure how often it does during the transfers) you can overheat your servers/networking gear and kill it if it is too old to have internal thermal monitoring and shutdown. Sudden power loss isn't healthy for any elctronics either and can kill them if it happens enough. (of course you know all this)
Recently we had a problem with one of the main power feeds to one of our data centers that killed all power to one side of our DC every 3 days. We couldn't get funding to repair it until we lost an entire 200TB SAN (thankfully it was mirrored with another SAN). Had the funding come in before the SAN was fried the cost to the company would have only been a day of an electricians time instead of SAN + electrician.
Don't let your management push it off until hardware starts dying.
Most people don't think clearly when they're on fire.