I'm asking for some tips to maximize the life of the cap as possible with the circuit design, such as undervoltage them, which is complementary to looking for the best capacitor ever created as well. I don't get your point.
Can you be more specific? do you agree because you have you experienced improvements at lower voltages than rated? I see this as a split decision because while some say it is useless, others say it can double or triple the life of the capacitor.
No, sorry, I can't. Hence my comment:
However, I can possibly invert my comment to say you must not exceed the rated voltage, doing so will shorten the lifespan. I have replaced many faulty electrolytic capacitors and, while I have not kept detailed records, I would say they were generally between 10 and 20 years old. Of course, they are the ones I saw as faulty. The ones that were working fine for longer than this I didn't see because they didn't fail so no one asked me to fix the not faulty equipment. Sorry, this is the best I can offer. Maybe someone else has better information, I shall watch this topic with interest.
Not me; I'd say pretty much the same thing, really.
I might only add that higher operating temperatures tend to accelerate (greatly) aging.
Otherwise, it's a bit of a crap shoot (pardon my French), although @Paul_KD7HB has said it in a more insightful manner. What he said boils down to the technology, material choices and manufacturing control are probably the key determinants of capacitor lifespan, with operating conditions/environment playing an also significant role. The latter you can influence to a great extent since this is the application you design. The former is a lot more difficult because making an informed assessment involves having rather deep insight into capacitor technology and manufacturing practices, and would likely involve on-site audits at manufacturers. Since this is evidently out of scope for the kind of project we're discussing here, the second best approach would likely be to just stick with a couple of reputable brands and steer clear of the (usually cheaper) generic ones, in the expectation that brand reputation correlates to some extent with technical product quality (which is already a tricky proposition to begin with).
Interesting - can you justify that formula?
Reading the info in the link as I expected the "useful lifespan" will be affected by the working voltage, but also more significantly by the ambient temperature, the ripple current and its frequency, and the quality of manufacture - which can be very variable even in a single production run. As evidenced by the spread in capacitance value
The temperature and ripple current appear much more significant than the working voltage.
So basically your guess is as good as anyone's.
Has anyone here made a "condenser" out of a jar with salt water inside, wrapped in foil and a metal rod held in the cap?
It's probably more of a junior science static electricity thing.
you should stay clear of the capacitors used in DeLorean and Tardis
machines, they're not known for their reliability
Ha ha ha ha!!! Your comment darn near had me spraying my coffee all over my computer screen and keyboard... 
I fully agree with you on both counts. Let me also add that people should also avoid any and all capacitors used in Death Stars and Tie Fighters for similar reasons...
@johnerrington
While reading this post on "Capacitor life span" I just noticed that you have a rating of "Faraday". The irony got me to giggle... 
I know from some experience that Sprauge made some very durable and extra long-life capacitors.
More specifically, I have in mind the filter caps used in the MA-117, MA-130, MA145, MA 160 vibration power amplifiers manufactured by Unholtz-Dickie Corp.
These are large banks of paralleled 19,000 uF at 80VDC and would actually be overdriven above their max voltage (90Vdc actual measurements, with a calibrated voltmeter). These amplifiers were built in the mid-eighties and have been powered up and running for weeks at a time, straight through. Down time has varied from days to a month or two. Operating temperatures have varied from 40 deg F through 100 deg F. Out of more than 500 of these cap, I've personally seen only three fail in over twenty years.
Faraday is one of my heroes - second only to Maxwell.
and his silver hammer?
Or possibly his demon?
This is only one example but I know the capacitors shown below are about 9 years old. (At least) one failed earlier this week. They are in a telephone system installed in 2015 and will have been powered continuously since then, except for the occasional power cut etc. I don't think 9 years is very good and a single example doesn't tell the story of all capacitors.
The top is swollen:
Close up, the foam is what is left of what should be inside the capacitor:
That's a good one!
Both are true pioneers, and heroes.
Speaking of lifetimes, I see ratings on Digikey all the time for Lifetime @ Temp. Even if charged below the rated voltage as the OP asked, is this a more indicative indicator of lifetime or just if it's in a high temp environment? Normally i'd think all factors come into play?
Yes.
Yes.
I have experienced capacitors have a life of 10 to 20 years, in fact most of the component failures I have seen are as a result of capacitor failure.
Life will of course be affected by extremes of environmental factors and voltage fluctuation.
I know nothing about EV car battery technology.
Do the battery packs use capacitors in the charging balancing circuit ?
If so, these capacitors would be a very weak point.
