This is a difficult question to evaluate because it really depends on your distribution system. As you know, when water remains in a line for a certain period of time, it will get a little stale. That being said, to provide a constant velocity in the short term may be compromising future growth (development) in the future because the line could be too small. With PVC, I use a design life of about 40 years and then try to crystall ball what the ultimate development demands are going to be on the main. Once the flow criteria is established, I then calculate the velocities to compare the present demand to the 5, 10, 20 and 40 year demands. In some cases, the water in the main will sit for a period of time and it's chlorine residuals drop, but are usually within the regulatory limits.
In most cases, complaints of taste and odour were common with the water that had no flow for a period of time.
The minimum velocity in the distribution pipeline has no limit, it depends upon the quantity of flow, the pipeline daimeter, amount of residual chlorine and bacterial contamination influx into the system.
For water stored in tanks for long time, there is a rule of thumb to circualte water from tank to tank atleast once a day and check for free chlorine.
Similar logic applies to pipeline. If flow exist take a daily sample for free chlorine analysis, otherwise if water is stagnant keep flushing for few minutes and take the sample for bacteriological analysis. Replenish chlorine if required.
For Velocity calculation in water system use the following equation :
V = Q / A ; whereas for circular pipes A= 0.785 D x D
V = Velocity in m/s
Q = Flow in Cubic Meter /sec
A = Area of cross section in Sq.m
D = Daimeter in meters
Usually you design the distribution network with minimum of 0.5 m/s and maximum of 2 m/s for uPVC 10 - 16 bar rated pipe