Document Ex9KjOnBbNGOboyvGxmDN9320

Sludge reCtreutah'dh In conventional lime-iodo softener, water Mows downward depositing its sludge, then leaves through inverted cone Pump' 0 Woter entering sludge-blanket softener settlies to ^bottom, and then must flow upward through the sludge to get out the pilot-plant stage, it showed possi bilities of providing much 'more effi cient silica removal than o new sof tener of the conventional type. A second-stage tank of conventional de sign was installed (or phosphate treat* ' menu Provision was made for lime and soda aoftenlng and ailica reduction in this softener, so that it might be used as a lime-soda softener while the pri mary unit is out of service for inspec tion. Piping was arranged so phosphate could be added at the feedwater pump section during such operation. In the new softener, raw water and chemicals are sprayed into a collecting cone at the top, Fig. 3. The mixture passes through a downcomer to the bottom, where the water is deflected upward through a blanket of sludge. Clarified water Is decanted to the phoaphate softener. SILICA-REMOVAL REACTION Magnesium hydroxide, commonly used to remove silica from water, may be produced by precipitation of the magnesium salts in the water by cal cium hydroxide. If there Is insufficient magnesium in the row water it may be added in the form of magnesium oxide or dolomitic lime. Activated magne sium oxide, especially prepared for the purpose, removes silica elleetively. At this plant It was used In both types ol softeners and was fed with high calcium lime and soda ash to meet our needs. Removal of silica proceeds slowly, approaching a certain equilibrium point for 'each aet of conditions. '-Several variables effect the speed and complete ness of the reaction. The equilibrium point approached is determined by: (1) amount of silica in the raw water, (2) amount of magnesium oxide em ployed, (3) water pH, and (4) tem perature in the softener. The speed of the silica-removal reaction varies with: (1) residual silica in the water, (2) concentration of magnesium hy droxide, (3) degree of unsoturstion ot i the magnesium hydroxide, (4) water; pH, and (5) temperature in the sof tener. As the sillcs is removed, coo- TABU 1 --Data Obtained from Softener Operation Canaan(lens/ Pafts-Msatrl ftlfles to rsw water. SIO. ppm Bilks In effluent, as 6lOt, ppm MiO read, ppm Parts 8lO, removed per part MtO Ava retention tiro* of Mj(OW),, hr Contact time of water with M(<OH)i, mto Hardness of affluent, as CsCOt. ppm Acrail toad ce softsnsr, % rated capacity Teroparattwa hi softener -Results bated on operation durina Juna-Oet, IMS. *Rmdta baaed oo operation during july-Sept, 1047. |<00) table tl--Typical .Analysts 01 Rew Water Toul kiriacu C*CO> Caletom hariom n CtCO, Mosecfloo hsrdnen u CaCOi KitlQit4n|t UktlMtr *a CsCOi Ptefuiphttuddn m C*CO Sitka nttOi ChiotidM M Ct| Bu)pb*t* M SO* pH Traipersture - 100 pp I M PP ' 40 pp \ ttopof 0 P** ` SJ PP 5 Opp 11 PP - i 74 t POWER July. H4R tioued removal slowa to such an extest that attainment of reaction equi librium in the softener is not feasible. PRELIMINARY WORK Research at the Pertnutit laboratories and pilot plant demonstrated the value of a concentrated sludge bed for silica removal. Data taken from this work were used in preparing curves, Fig. 4, which show the amount of silica re maining in solution plotted against magnesium-oxide feed, both with and without a sludge blanket The curves show ihat silica removal is much greater when the reaction Is carried out in a concentrated blanket of sludge. Fur thermore, in the region of Jow-eilica concentration, relatively largo increases in mognesium-oxido feed are necessary to achieve further reduction in silica, with or without the sludge blanket. SILICA-REMOVAL COMPARISON Although reaction equilibrium is not reached in the new softener, it is much more closely approached than in the conventional type. The greater effi. cfeney of the sludge-blanket unit ft attributed primarily to greater concen tration of magnesium hydroxide with which the water comes in contact. Some ioereate in efficiency Is due to higher operating temperature. Table I provides a comparison of the two types of softeners over comptrable periods, and shows the new unit's greater efficiency, silica removal per unit of mognesium hydroxide being about double. Ia comparing operating efficiencies, it should be pointed out that during the periods considered the new softener was operated at a temperature approx imately 23 F above that of the old units. Whereas extensive data regard ing the effect of temperature on effi ciency are not available, the results from operation (or (our days at lower temperatures indicate silica content of the effluent may increase by 1.5 ppm Jf the temperoturo in the new softener dropped from 243 F to 220 F. There was no significant difference in contact time ef water with mag nesium hydroxide in the two types of o tenets. The contact time in the new nn t could have been increased by rais ing the sludge level. This was unnecessary, however, because incretaing e flow through the softener from 46 Percent to 80 percent of rated capacity creased silica content of the effluent nlf one ppm. blown down from * softeners once each eight hours, nerea only ^ of fte ,|udge w#i P9W|r July 1948 blown down from the new unit once every two days. The average reten tion time of the magnesium hydroxide in the old softeners was 4.7 hours, compared to lOO hours in the new unit. Concentration of sludge in contact with the water In the old softeners averaged 1500 ppm; the added magnesium hy droxide was 23 percent of the sludge. Concentration of sludge in the new softener was 60,000 ppm of which 17 percent was added magnesium hy droxide. Added magnesium hydroxide in the sludge layer was thirty times as con centrated as in the conventional sof teners, although only 66 percent as much magnesium hydroxide was fed. As the speed of the silica-removal re action is proportional to the concen tration of magnesium hydroxide, and the contact time of the sludge with the water is the same in either cose, it would be expected that more complete silica removal would occur in the new unit. In the new softener, part of the un spent magnesium hydroxide settles out below the baffles st the bottom of the downcomer, requiring that sludge be recirculated from thia area to obtain maximum efficiency. Rate of recircu lation U not important; it may he stopped several hours without adversely affecting silica removal. Recirculation of sludge in the old softeners caused the magnesium hy droxide to be re-used, and also in creased the concentration of sludge in contact with the water from 480 ppm to 1500 ppm. The average retention time of the sludge was increased from 1.5 hours to 4.7 hours. Twice as much magnesium oxide was required when sludge was not recirculated; failure of sludge recirculation equipment,con sequently reduced silica removal. The amount of ailica removed in the new softener increases with tho depth of the sludge layer through which the water passes. During the operating period on which the figures in Table 1 are based, the sludge level was about 7 feet above the conical section. Rais ing (he level above this point improves silica removal only slightly ot normal load. Fig. 5. EFFLUENT HARDNESS Chemical feed to the new softener may be discontinued several honrs without causing a dangerous increase in silica. This is due to the stabilizing effect of 4000 pounds of added mag nesium hydroxide in the sludge layer, Failure of chemical feed to the old softeners, on the other hand, always 4 Residual illlco In the effluent with ond without sludgo-blonket action Silica in row wafer, 55ppm Femosil feed, 52ppm 8 --1--\--r~r-- Sludge contact time -- 4*ft depth, 12min -- S'ft depth, 25min V 16*ft depth, 56 min 1 5a i (At/6*t7 depth, sfa \ **.? in etraigh v* e 83 4 $ B 10 12 14 16 Sludqe depth, ff 5 Raising sludge level obove 7 ft im proves the silica remavol only slightly caused a rapid end immediate increase in silica content As shown in Table II, the raw-water hardness averaged 100 ppm. Hard ness of the effluent from the new sof tener averaged 15 ppm os compared to 21 ppm from the old softeners. The lower hardness was due in part' to the stabilizing effect of the undissolved calcium hydroxide in the sludge blan ket, and in port to the lower solubility of calcium hydroxide at the higher operat ing temperature. Following stoppage of the chemical feed for two or three hours, treoted-weter hardness increased to only 25 to 30 ppm. This resulted in part from the 10 ppm excess sodium alkalinity of the raw water. Although the percentage undUsolved calcium hydroxide In the sludge of the old softeners was higher than in the sludge-blanket unit, the total (Continued on page 156) (401) V