Aguato

    COD and sulfate removal from cooling tower water by photoelectrochemical oxidation

    Water from the outlet of a cooling tower system in the tannery industry in Ranipet, Tamil Nadu, India, carried organic nutrients that can cause microbial corrosion of the tower's metal surfaces. Batches of this water were treated in the laboratory to cut pollutants and limit microbially driven corrosion.

    Treatment line

    1. 01Photoelectrochemical oxidation cellA tubular titanium flow cell with a coated titanium mesh anode and a 6 W UV lamp in its centre, through which a 2.5 litre batch was recirculated by peristaltic pump at 440 mL/min under direct current.
    2. 02Sunlight exposureThe treated water was then left in sunlight to remove remaining contaminants.

    What the source found

    The combined UV and electrochemical process worked best, ahead of electro-oxidation alone and UV alone, removing 93% of COD, 94% of sulfate and 86% of hardness. Steel coupons kept in the treated water corroded far more slowly than in untreated water, which is attributed to fewer nutrients for bacteria and to disinfection by electrochemically generated hypochlorite.

    Batch runs of up to 6 hours sampled every 2 hours, comparing untreated water with electro-oxidation alone, UV alone and the combined process, followed by 14 to 21 day corrosion tests on mild steel coupons.

    Water in, water out

    ParameterSymbolUnitInletOutletRemoval
    Temperature°C2835
    pH5.715.7
    TurbidityNTU5.090.4891%
    Total suspended solidsTSSmg/L1,00013387%
    Total dissolved solidsTDSmg/L700225
    ChlorideCl−mg/L634.55367.242%
    Electrical conductivityECμS/cm2,5151,348
    CODCODmg/L2,51218593%
    SulfideS2−mg/L88811287%
    SulfateSO42−mg/L1,95712394%
    Hardness as CaCO3THmg/L1,25017086%
    AlkalinityAlkmg/L1,746212

    Inlet: untreated cooling tower water (control). Outlet: water after photoelectrochemical oxidation. One laboratory batch run; values are means with standard deviation (deviation not copied); the treatment time behind the table is not stated. Values and units are as the published source prints them. Removal is worked out from the two values, or is the source's own figure where it prints no pair.