Base welds on poles are fine if done right. An after the fact visual check can indicate bad or potentially bad welds but really can't confirm weld soundness. Radiography would be the way to go but cost might be an issue.
Do you know who manufactured the poles? A reputatable, established manufacturer would have a good QA/QC program to ensure weld soundness. The only significant problems I've heard about in Texas were stadium poles made by a manufacturer who didn't have pre-qualified welders or pre-qualified weld procedures. They tried to supply poles to TxDOT but failed to achieve prequalication. They had been supplying light poles to various schools districts, none of which made any attempt to prequalify fabricators. They just took what they got until the poles started blowing down in relatively light winds. The poles weren't more than a few years old. Inspection revealed other poles with fatigue cracks and defective welds, so everything this manufacturer made had to be replaced. The schools took a big hit because the manufacture had gone bankrupt shortly after being turned down by TxDOT.
Even if you consider that the manfacturer produced a sound weld, you would still have an issue with possible fatigue cracking because even sound welds may eventually develop fatgue cracks. If it has a socketed base plate connection this type of crack would likely form at the weld toe-to-pole intersection and would be detectable by ultrasonic, dye penetrant, or mag particle techinques.
Another thing to consider, is whether the lighting engineer can modify the lighting to reduce the wind load. This may be doable with fewer, higher intensity lights. The lighting engineer I worked with on a similar job was able to do that and he also reduced suspended weight by moving the transformers out of the light fixtures to boxes mounted close to the base. He said this was a trend with stadium lighting so it may already be the case on your project. Just thought I'd throw that out there.
