Back in college a few years ago I wrote a paper trying to compare the environmental cost of maintaining and operating 90s cars vs buying and replacing electric car batteries measured by co2 emitted (typically written co2e). This was written around 2012 but I found the break even around 10 years. That is, electric cars were measured as a fairly flat line (not fully flat, assuming some non zero co2e for electricity generation). There was of course a ton of hand waiving but the gas car was assumed to have a y intercept near zero and to be operated for 10 or 15k miles/year (can't remember exactly but I know it was in that range). I would link it but I'm currently not in the same country as the hard drive that might still have a copy of it. Of course I would prefer to have a peer reviewed publication on this, but I had a lot of trouble researching the manufacturing/mining estimates back when I wrote it.