1. Limited time only! Sign up for a free 30min personal tutor trial with Chegg Tutors
    Dismiss Notice
Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

UV disinfection in water treatment

  1. Apr 19, 2010 #1
    I am designing a UV disinfection unit
    think: http://www.wedeco.com.au/www.australianultraviolet.com.au/animation2.gif [Broken]

    I need to ensure an internal intensity/dose of 410 J/m.sq to all waters that pass though. (to comply with WHO guidelines)

    I am aware of The Beer-Lambert Law but will have no data on absorption apart from turbidity in NTU's. Is there any conversion or link up?

    The text books I have looked at have the photon output and as such energy within the tank but not intensity. (they are dealing with oxidation of chemicals and not inactivation of cells)

    Any recommendations (ideas, online reading etc)

    or am I just going to have to over design with a intensity sensor on the outer casing and power control system.
    (humm going to need at least an ball park idea to do the over design)
    Last edited by a moderator: May 4, 2017
  2. jcsd
  3. Apr 19, 2010 #2


    User Avatar
    Science Advisor

    The Beer-Lambert Law may get you the answer you seek, provided that you know some things about the water that will be passing through the filter. Most importantly, will the particulate matter have been filtered out before the water passes through? If not, you will have to account for scattering as well. That is also doable, but it complicates things a bit. Other important factors are the expected concentrations of organic molecules and metal ions, both of which can have significant absorptions at UV wavelengths. As usual, the more you know about the expected composition of your sample (i.e. the water), the better.

    If your water is filtered, and you know the max concentrations of absorbing species, then you can easily calculate the fraction of emitted light that reaches the inner radius of your flow tube from the Beer-Lambert Law. A few unit conversions should get you a ball-park estimate of the lamp intensity that you are going to need (you will need to account for the flow velocity as well).
    Last edited by a moderator: May 4, 2017
  4. Apr 19, 2010 #3
    "you will have to account for scattering as well. That is also doable,"

    I am going to assume that not all partials have been filtered out
    so, how do I account for scattering?
  5. Apr 19, 2010 #4

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

    Since your standard is energy/area, I am thinking you also have a requirement for 'dwell time', or how long the water is exposed to the output. Energy = power/time, and power is pretty much given by the intensity- there are some subtle radiometric issues, but the bulb manufacturer should have an irradiance specification that you can put to use.
    Last edited by a moderator: May 4, 2017
  6. Apr 19, 2010 #5

    yeah of course,
    however I was planning to do it the other way round i.e. getting different standard bulbs and seeing what flow I could get passed them. By finding the time needed in the chamber and the distance is the length of the bulb.

    The problem is more insuring the intensity with differing water quality at the edges of the chamber.

    A GL30 outputs 13.4W UVC
    (http://www.globalmarket.com/product/Shot-wave--Ultraviolet--Germicidal-Lamp_25047.html [Broken])
    (its a crap bulb I just Googled it for real data)

    Its 25.5mm dia so call it 30mm with a water proof sleeve and 893mm long (ouch that's long for 13W, anyway) so call it 800mm of useful length (taking off the caps)

    and I'll make the chamber 200mm dia (internal surface of outer casing)

    get it all in to SI units


    (pie *D)*L= Area= 0.50 m.sq

    so the outer surface (in a vacum !!!!) gets: power/Area= 25.86 J.s/m.sq

    and 410/25.86=15.85 sec in the chamber
    0.8m/15.85=0.05 m/s flow velocity

    (area-area of bulb) *0.05m/s = volume of flow

    approx 2L/s

    But anyway that's the easy stuff
    going back to
    "so the outer surface (in a vacum !!!!) gets: power/Area= 25.86 J.s/m.sq"

    its not in a vacum and will be adsorbed along the way
    and the beer's law refers to molar concentrations that I have no data (other then the aforementioned NTU values)

    So I suppose a better question is where can i get info on adsorption in cloudy substances
    because for an over design I could assume that only a small percentage reaches the outer surface and then follow though the eqn's.

    I just need an idea and like I said I'll use a sensor.
    Last edited by a moderator: May 4, 2017
  7. Apr 19, 2010 #6

    Andy Resnick

    User Avatar
    Science Advisor
    Education Advisor

    Last edited by a moderator: May 4, 2017
  8. Apr 19, 2010 #7


    User Avatar
    Science Advisor

    Like I said, it gets more complicated. I was going to write up a little on Mie scattering here, but I see that you already have NTU ratings for the water sample. I am not an expert in the field, but those are a measure of perpendicular scattering due to turbidity, and thus it should be possible to convert them in to some sort of parallel form for transmission. If you look up some references on nephelometry, you may be able to find this information.

    However, the rated specifications for drinking water have rather low NTU ratings (well less than 1), so you may find that the losses from scattering are negligible. If you are dealing with significantly more turbid samples, then you may have to address this in more detail. Regarding absorption, at the wavelengths of the bulb (~250 nm), water has a non-zero absorption cross-section, but the total attenuation is only around a percent or two for a 10 cm path length (based on quick interpolation from wikipedia spectral data ... you should check this). Again, assuming this is drinking water, then total concentrations of UV-absorptive materials should be small, but this can be measured (see below).

    So, I think your approach of buying the cheapo-bulb and using a sensor to measure the transmitted power directly is the best option. My guess (but it is only a guess) from the info you have provided is that the scattering/absorption effects will be fairly minor. One thing that would be nice to have before jumping in is an UV-absorption spectrum of a sample of your water at the wavelength of interest ... this may be tricky to get, because many UV-absorption instruments only go down to 300 nm (if that far). However, if you are at a college or university, you may get lucky. That will allow you to measure the absorbance of your sample for a known path-length, so it will be trivial to extrapolate out to your actual distance of interest. (Note that this measurement will also account for the scattering effects, since all light that doesn't make it to the detector will be attributed to absorbance.)
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook