I have a 1000uf 10v filter capacitor going bad yearly in the 5volt section of an ATX SMPS.
A 1000uf 16v capacitor is physically larger and i have room for it.

Do you think it would last longer?
Could there be some problem using a higher voltage rating?

 

A higher voltage rating shouldn't hurt, but using a cap rated at 10V in a 5V circuit is already fairly conservative.

ATX power supplies are throwaway, but a good quality supply will last decades. I built a computer with cherry picked components and ran it 24x7 for a couple decades before retiring it. I've only had one ATX supply fail and it was a pain in the rear to isolate.

 

I have a 1000uf 10v filter capacitor going bad yearly in the 5volt section of an ATX SMPS.
A 1000uf 16v capacitor is physically larger and i have room for it.

Do you think it would last longer?
Could there be some problem using a higher voltage rating?

Make sure the capacitor you use is low esr and high temperature type.

 

What brand and series is the capacitor you want to use?

Electrolytic caps are highly variable in their performance and lifetime ratings. These days a good quality low ESR type, depending on physical size, should have a lifetime rating of at least 5000 hours at 105 °C at rated ripple current (and that is a widely varying parameter). If ripple current is less than the rating, longer life can be expected - heating due to ripple current is harder on the cap than external heating to the same temperature. Operating below rated temperature will approximately double the life expectancy for each 10 °C reduction in temperature, all other things being equal. When I was designing industrial switchers I aimed for a life of 100 000 hours. You can't be casual about how you choose caps for that objective.

Nichicon, United Chemicon (Nippon Chemicon) and Panasonic have very good series for switchers. Rubycon, once an embarrassment to the name capactor, is reputed to have some premium lines, too, though I have never used them. The solid (polymer) electrolyte aluminum types have extremely good specs. I always used to be disappointed when I considered them because they were never available in high enough voltage and/or capacitance for my purposes, but that has changed and the range of choice is much larger.

Using a higher voltage capacitor in the same series will generally yield longer life simply due to the fact that it is physically larger. Aluminum electrolytics don't need to be derated for voltage. A tall skinny capacitor will have better specs than a short fat one in the same series with the same capacitance and voltage rating.

Although it must be approached with caution because it changes the frequency response of the output filter in a way that can mess with stability (removes a high frequency "zero") adding a few X5R or X7R 10 V surface mount ceramic caps of at least 10 µF each can unburden the electroltyic of much of the high frequency ripple current. Such caps have become quite cheap and easy to get, though hand soldering SM ceramic caps with an iron is best avoided (caps subject to cracking). Stay away from the slightly cheaper Z5U and Y5V and similar types. They have horrendous negative voltage coefficient of capacitance.

The power supply you are using is a forward converter (one of several transformer coupled derivatives of a buck converter). They are quite gentle with the output caps. Flyback converters are brutal to them.