Page 315 -
P. 315

10.56                      CHAPTER TEN

        sured. Several guidelines have been developed to assist the designer in ensuring adequate
        contact between the ozone gas  and the liquid:
        •  Maintain a gas-to-liquid ratio between 0.05 and 0.20. This guide comes from the chem-
          ical engineering field for  effective  transfer and contacting in a  two-phase  (liquid and
          gas) system. However, this may be difficult to maintain under a wide range of flow and
          dose conditions.
         •  Maintain a minimum gas floor loading rate of 0.12 ft3/min/ft 2 (0.037 m3/min/m2). This
          guide comes from studies on activated sludge systems and represents the minimum gas
          flow rate for proper mixing of a liquid mass.
         •  Maintain uniform floor coverage  with  diffusers  spaced  at  no more  than  3  ft  (0.9  m).
          This guideline was developed in recognition of opportunistic flow of the liquid through
          areas low in or devoid of gas bubbles.
           If followed, these guidelines will yield a design that effectively promotes uniform con-
         tacting and distribution of ozone throughout the  liquid. In disinfection applications, this
         means the designer can be assured that all the liquid has been equally exposed to the dis-
         infectant. These guidelines work well with conventional air feed systems operating at typ-
         ical water treatment doses.
           However, recent developments in ozone generation, resulting in ozone concentrations
         in excess of 10% when using high-purity oxygen as the feed gas, have made it almost im-
         possible to  meet  these  guidelines when  operating  at typical disinfection doses  of  1 to  3
         mg/L. Solutions to these applications have included supplemental mixing with air or wa-
         ter injectors. One manufacturer recommends using more,  small  [4-in.-diameter (10-cm)]
         diffusers with smaller bubbles (1 to 2 mm) to improve the floor coverage. Ultimately, the
         higher  ozone  concentration available from  the  latest generation of generators  is forcing
         the  industry to  consider  modifications, and  alternatives,  to  the  conventional multistage
         fine bubble diffuser contactor.
           The  search  to  optimize  the  hydraulic efficiency  Tlo/T in fine bubble contactors has
         been  investigated by  using tracer  dye  studies  of existing basins, on computer modeling
         based on finite element analysis, and using computational fluid dynamic techniques. While
         complex in development, this work provides practical results for the designer.
           By using D,  L,  and  W to represent the depth,  length in the direction of flow, and the
         width perpendicular to the direction of the flow, respectively, it has been shown that hy-
         draulic  efficiency  is  very  closely  related  to  D/L.  While  previous  sources  have  recom-
         mended basin configurations of  1.5  to  1.0 to  1.0 (D,  W, L),  the  TIo/T can be improved
         by 50%  or better if D/L is increased to 4.0 or higher.  Hydraulic efficiency appears to be
         independent of width. Consequently, to optimize  Tlo for contactors with a depth of 20 ft
         (6.1  m), the cell length should be less than 5  ft (1.5  m).
           Appendix O  of the  SWTR Guidance Manual provides extensive discussion in evalu-
         ating CT for existing basins. For most applications, the designer may use the T10 method
         for  estimating basin performance.  However,  the  determination of  CT for  ozone contact
         basins is complicated by  the  time required to  establish a  residual  concentration and the
         relatively rapid decay of ozone residual in water.  Consequently, a cell-by-cell evaluation
         is  necessary, and the  CT for  an ozone  contact basin is the  sum of the  CT values for all
         cells.
           The designer may conservatively estimate the  Tlo/T for the entire contactor. The D/L
         value may be estimated by

         D
         L
                                    D  × number of cells
             total contactor length (excluding baffle thickness and counting chimneys as cells)
   310   311   312   313   314   315   316   317   318   319   320