Question about the thermodynamics of a sewer system

[JJ_TT_HH]

Hi there,

I have recently become involved with a scheme in which temperature is being monitored in a sewer system to identify if the system is being compromised by infiltration. As mixing of temperatures in liquids over time and length, heat loss in pipes etc. isn't my specialty, I would like to open this topic to debate and suggestions.

So here it is:

There is a large expansive sewer network in a catchment. Temperature is being monitored at the outfall of 20 pipes. Values are returned in Celsius every few minutes. Other than the normal foul flows there may be infiltration getting in to the system in parts.

Is there an equation (or a number of equations) that can do the following:

A) Knowing the dimensions of the sewer, given an average flow and other average set variables. Estimate the temperature of the liquid exiting the pipe, given the temperature going into the pipe.

B) If A. is possible and the real temperature is significantly cooler/warmer than the estimate. Then estimate how much colder/warmer liquid is entering the system.

I understand its a lot to think about and I would appreciate any advice/help given.

Thank you!

In response to the answers given:

Thank you for replying. To clarify - there is currently 3 months worth of data (and ongoing) with data being recorded every 6 mins - like the picture shown. Rainfall data is also available for those 3 months, as is ambient temperature and flow data. However, the aim is to use temperature sensors and not flow sensors in future so I wanted to look at this without flows and use an average for that time of day. The infiltration is suspected from a river so is assumed as constant.

This scheme is to be used as an indicator (as mentioned)- to potentially highlight areas in which to investigate - that is the final aim - SO a very simplistic scenario of water of a temperature into a sewer, set length, no turbulence, constant flow - no infiltration. What would be the temperate loss of the water per unit length of the pipe. Or what would the temperature out be? Hope this helps!

 

[ashtree]

I can see what you are trying to do but i think there are too many unknown variables to do this with any sort of accuracy.

Sure a change in temperature from the norm might indicate infiltration but it could indicate a lot of things. Sewage flows coming into a plant tends to be fairly consistent in temperature on a particular day, but you do see variation over a wider period of time.

In saying that i was involved with a plant on the west coast that used to see twice daily temperature swings whhich was caused the local casinos through ice down the sink after the morning and afternoon peaks related to the commuters who used to drive between two cities.

Monitoring the sewer would give you some good baseline temperature and maybe you could then use any change to indicate:

1) When infiltration starts.
2) Duration of an event.
3) A possible order of magnitude of an infiltration event.

I would be thinking that maybe a measure like conductivity, although also difficult may have fewer unknowns.




Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[itsmoked]

I believe you can do this but would need to gather lots and lots of data that would include more than just a few inflow temps.

You will also need:
The temp of the city potable supply.
Time of day.
Ambient temperature.
Precipitation (rain gauge) measurements in several areas that could contribute to any infiltration.
Temp and possibly flow measurements at a few of the largest sewer users.
And all of these readings on the order of 5 minutes.

With this data you can then correlate the data and get a confident feel for what is actually going on.

Since your focus is on infiltration it's imperative that you have the most important data that drives it which is precipitation.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[ashtree]

But there are also issues about depth of sewers, the amount of turbulence, detention times, flows being pumped in from sewer pumpstations.

I believe that temperature would be a useful indicator of infiltration but there are too may variables and too much data to consider it an accurate way of measuring infiltration over a broad area.

Regards
Ashtree
"Any water can be made potable if you filter it through enough money"

 

[BigInch]

Geez. You might just get the temperature of the water. Both relative volume and relative temperature of ALL inflows, residence time in pipes, changes in ambient temperatures will all make that calculation very difficult. I suspect you might get more confused trying to see nonexistent patterns where they do not exist and miss a place or two where they might. I'd like to be a fly on the wall when you have the meeting to explain the results.

Identify suspect areas and monitor or estimate all known inflows and outflows. Infiltration, or exfiltration, will be indicated by any differences.

Reaction to change doesn't stop it

 

[cvg]

model the system in SWMM. I believe it has all the algorithms to do what you are asking, only intended for stormwater, not sanitary. I can't comment on the details, since I have not attempted this and I agree with others that you may not be able to determine a good correlation between temperature and flow data sufficient to locate the leaking and amount of leakage. A more typical method of locating I&I sources is to monitor flow, not temperature.

https://www.openswmm.org/Topic/4702/water-temperature-modeling