Water Age - InfoWater

[ESYSU]

1. Has anyone used the age calculation feature in InfoWater to gauge the tank turnover time and/or relate it back to water quality?
   
   • I'd be grateful for any case studies! Especially if it's been compared against the 5-day rule of thumb (http://www.innovyze.com/files/2018/I...iencies-WP.pdf - Page 5).
   
   
2. Any suggestions on how to characterize what mixing option should be used without any CFD Modeling?
   
   • Or without CFD Modeling, is completely mixed the way to go? I believe completely mixed is more conservative in that it typically reports a higher age than the other mixing options.
   
   

Any input is appreciated! Thanks!

[Patrick Moore]

Elena,

Water age is used as a surrogate for water quality issues quite often as most water quality issues don't become problems in a water system unless they have enough time for them to occur. But what age equals a problem is highly system specific. While the general rule of thumb you hear a lot is "five days" some systems can go much longer before they have concerns and some systems may have issues only after a few days. What qualifies as a problem in water age would be highly dependent on your specific water system.

But if you have a well calibrated water model and run a water age you may be able to match up areas with known water quality complaints with typical water age values seen in the model as a way of providing guidance on what age may have issues. The water quality monitoring staff at your utility would also serve as your primary and best resource for your specific utility as they are the most knowledgeable on your systems water quality issues and where problems (taste and odor complaints usually) typically occur.

The methods they use are very handy and I have worked with Walter before on a project where these were applied, and they can be super useful. The largest thing I tokk for that was that many tanks may get "enough" momentum each day to be close enough to complete mixed that you can use it. This has several nice tools that can help determine if you are more likely or less likely to have a mixing problem given the tank configuration, and tank inflow and outflow.
AWWARF with Walter Grayman have done several studies that can be very helpful in determining if a given tank has a mixing issue. I suspect he documents many of them in this article: https://www.jstor.org/stable/41311897?seq=1#page_scan_tab_contents or this article: https://awwa.onlinelibrary.wiley.com...1993.tb06026.x


One thing to be aware of in a hydraulic model is that the methods that are not Complete mix , may not always be clear in what results they are presenting. The tank WA values represent the water quality of the water leaving the tank, but except for complete mix (where the tank concentration is the same everywhere) the concentration the model will show will not fully reveal the entire WQ of all of the water within the tank. Both the two compartment modeling and LIFO (Last in First Out) and FIFO (first in First Out) methods may have much worse water quality within the entire volume of the tank, but it is not "visible" to the user until a slug of "bad" water leaves. This is one of the limitations of the LIFO, FIFO, and 2 compartment methods as you cannot see how "bad" all of the water truly gets, you can only see the quality of the water leaving the tank. For this reason, it is always wise to run at least a few scenarios with complete mixing so that you can fully compare.

I would actually consider the Complete mix tank as the "best case scenario" for WQ mixing as it does not allow water to stagnate within a tank and you know fully well what the concentration is everywhere. There are many methods employed now a days as well to add mechanical mixing (Flow Bee, PAX mixer, and even manifolds with tideflex valves that are relatively low cost and ensure the quality in the tank is consistent at all times and that temperature stratification is not occurring and that no dead zones occur.

If dead zones exist or if you use LIFO, FIFO, or 2 compartment mixing in a model, you may actually have a much worse WQ problem occurring within the tank because the tank quality parameter only shows quality of water leaving the tank, not an "average" water quality within the tank. All three of these mixing methods can potentially mask a mixing issue and its impact on the water system because of this and why at least some runs as complete mix should be completed for comparison. If you don't have an issue when completely mixed in the model, does not guarantee there would be no issues in the real system, but it would tell you the potential "best case" WQ you could get if a tank was completely mixed. a non mixed tank would just "hide" the WQ problems better until some of the bad water was released. This to me is one of the biggest concerns associated with the mixing methods used by EPANET. If the Tank WQ represented the average WQ parameter in the entire tank, then they could better help users understand, the issues associated with stagnation and short circuiting, but until this occurs, use some caution when choosing model tank mixing methods that are not complete mix.

If you have a chlorine autosampler near a tank I would review the WQ results and look for any sharp drops in the chlorine levels near the tank as possible indicators of a mixing problem for that tank. Otherwise using something like the PAX mixer studies do with temperature readings at various tank levels is an easy non-CFD way to text for tank mixing: https://www.paxwater.com/success-stories.

Hope this helps:

Patrick Moore

[ESYSU]

Thank you for the insight and references, Patrick! This is very helpful. I will definitely keep this in mind.

Also, I found a graph (see below) that compares % daily volume turnover and water age. However, I am not sure how it compares to the hydraulic modeling method.
Additionally, it says a 20-30% daily volume turnover is recommended for a 3-5 day water age. Do you know if there is a way to relate the InfoWater results to a % daily volume turnover?
Age vs Daily Turnover.jpg

Source: https://www.onewaterohio.org/docs/14...Tanks.Duer.pdf (Page 4)

[Patrick Moore]

Elena,


The model can calculate water age, but "tank turnover" is not something you can calculate directly within the software, but you should be able to use model results to calculate it. However, this specific graph could likely be very system specific so I would use caution in using it directly without further research as it could likely be based on a specific tank and a specific system. I would read the report it came from for documentation on how you might do something similar if desired.

Patrick Moore

However, this graph is pretty much similar to the methodology Walter Grayman discussed I believe regarding getting a sense if your tanks were likely mixed or not mixed, but a lot depends on the inlet configuration as well:
(click if need a larger image)
mixing equation.jpg

Thank you for the insight and references, Patrick! This is very helpful. I will definitely keep this in mind.

Also, I found a graph (see below) that compares % daily volume turnover and water age. However, I am not sure how it compares to the hydraulic modeling method.
Additionally, it says a 20-30% daily volume turnover is recommended for a 3-5 day water age. Do you know if there is a way to relate the InfoWater results to a % daily volume turnover?
Age vs Daily Turnover.jpg

Source: https://www.onewaterohio.org/docs/14...Tanks.Duer.pdf (Page 4)