smoke leakage test for duct

First of all, it depends on your specifications. What test was required? Typically I would require something that comes from the SMACNA standards. If your project only gives the requirement of 1% (which is indeed tight), then the contractor has to propose more than just a method, they have to show how that method will demonstrate the standard. Personally, I do not feel smoke is a good test other than it sounds good. How do you quantify smoke? There is a standard test which involves sealing the ductwork system outlets and then pressurizing the ductwork to the operating conditions (perhaps 4" if a VAV system) with a characterized fan that allows precise measurement of airflow. Then this flow is compared to the maximum system airflow to gain a percentage. Seek guidance from a qualified testing and balancing agent...they know the right way to do this test.

Ok, not sure that I understand how they would do that, but I would not recommend smoke. Most "cold smoke" machines use some kind of oil or other material that looks like smoke but is cold and should not irritate lungs unless way too much is used. It does not act like smoke in a fire and I would have a hard time seeing that it would do what the contractor is hoping for. If they are testing the longitudinal seams on the ground, I could see creating some kind of appliance that blanks off both ends and then allows application of low level compressed air. Too much and you would damage the seam, too little and it would not be a good test. If they persist, I would ask for more information on how they are making the smoke and what the material is and how it might affect other credits.