Water Model Rules of Thumb?

[jartgo]

We're having a small debate in our office about how much field work should be done in developing and calibrating a water model. That is, once the system is sufficiently mapped, what percentage of hydrants should be flowed in order to establish flow conditions. Or perhaps you look for a certain number of hydrants per mile of main? Just curious to see what others consider an adequate first stab.

Any documented references along these lines would be helpful as well.
Thanks in advance.

 

[UtahWater]

We don't shoot for a certain number. We go for what looks to be spatially representative and hydraulically unique situations seem to dominate, with only a few spread throughout the 'heart' of the network. I'd say for a rough guess that we request tests on only 2-5% of the system's hydrants, but eagerly receive more flow results whenever available. I know it's a pain for the field guys, but asking them to collect some data when they're flushing anyway is usually good for timing, as long as you're aware of any unique conditions they may be creating (closed valves when doing unidirectional flushing).