The recent Wallingford Report SR 661 compares previous research and presents a best fit solution for movement of air in pipes variables are slope, diameter, velocity and air pocket shape and volume. The equations presented are best fit. (The study report was available free on the Internet but it appears that they have now updated the report to a design manual which costs some $50 to download).
Using the HR proposed formula, as stated above by Balebeng ,and applying a FOS of 1.1, as recommended by HR, gives a critical velocity to move air along a horizontal pipe of about 1.13 m/sec. (c.f 1.2 m/sec stated above by BIMR)
The report uses S in degrees, but if you are working on an Excel spread sheet it is easier to work in radians and as Theci states you can use either as long as you are consistent.
For a slope of 1/125 (0.46 degrees) the slope is not significant in comparisaon to the scatter of the data and to be safe, unless you have a velocity somewhat in excess of 1.2 m/sec, I would consider that the air pockets may not move downstream and would provide for air release valves on the downward slope. ( depending on the length and frequency of pumping - obviously air will travel back upstream to high points when the pumps are off)
(PS Falvey's paper USDI - Engineering Monograph No 41 is available on the internet with a google search)