# Thread: Water Age - Alternating Sources

1. ## Water Age - Alternating Sources

Hi,

I am having difficulties in H2OMAP modeling water age for a system that alternates the system water source weekly. The first week water enters the system system directly through a PRV. The second week water enters the system by filling a ground storage tank which is then pumped into the system as needed. The third week it pumps directly into the system again (with the water in the ground storage tank stagnant for the week).

I can run either source scenario separately; however, I don't think that is an accurate representation of the system as it does not account for the age of the water in the ground storage tanks during the week it is not pumping.

Any suggestions on how to model this system?

-Hannah

2. Hello Hannah,

This sounds like a challenge to model. If you want to use the separate scenarios method, you could specify the initial water age in the tank to match a hand calculation of what the worst case age would be at the start of the week in the tank, but you wouldn't get the effect of fresh water flushing out the tank as week of using the tank progresses (if the tank's pretty big, assuming the worst case age may be an acceptable solution).

You also could run the full multi-week scenario for like a month or so (adding in the tank if it's not modeled already)...I've ended up doing this for distribution systems with a lot of pressure zones. The key to this strategy is re-working all the controls and demands in the model to be a really sustainable run so there won't be errors during the simulation.

If the tank and PRV are far apart from each other, you could also use the multi-trace to figure out the coverage area each of the sources provide...don't know if this will really help or not.

Thanks,
Brian

3. Hi Hanna,

Brian had some great suggestions. I would recommend the multi-week scenario. To replicate the changing water sources set simple controls based on run timers on the pipes providing supply each week. Since 7 total 24 hour days equals 168 hours, the controls changing supply will need to be made at those points in time

Timers:
Element ID Time = 0 Time = 168 hours Time = 336 hours Time - 554 hours Time - 672 hours etc...
Pipe 1 Status - open Status = closed Status - open Status - closed Status - open

Example (click if need larger image)
timer control.jpg

If you set controls like this for multiple weeks you can control how water can enter from each source. Adjust the timer controls so they match your desired operation and set the initial water quality or quality source from each source reservoir (if non zero) so that those have the desired baseline water age if not assumed to be zero.

This should allow you to calculate the impact of the actual operation. The key is using pipe controls to control the water supply from each source. For the PRV I would control the pipe downstream of the PRV.

Best to you in modeling this as it can be a little challenging to set up timer controls for each source.

Patrick Moore

Originally Posted by Brian
Hello Hannah,

This sounds like a challenge to model. If you want to use the separate scenarios method, you could specify the initial water age in the tank to match a hand calculation of what the worst case age would be at the start of the week in the tank, but you wouldn't get the effect of fresh water flushing out the tank as week of using the tank progresses (if the tank's pretty big, assuming the worst case age may be an acceptable solution).

You also could run the full multi-week scenario for like a month or so (adding in the tank if it's not modeled already)...I've ended up doing this for distribution systems with a lot of pressure zones. The key to this strategy is re-working all the controls and demands in the model to be a really sustainable run so there won't be errors during the simulation.

If the tank and PRV are far apart from each other, you could also use the multi-trace to figure out the coverage area each of the sources provide...don't know if this will really help or not.

Thanks,
Brian