Document mpgYnQBKgk5zGY5ngXMZdybzk

TABLE I: ANALYSES OF WELL WATERS IN PARTS PER MILLION Well N 747 701 Ctpadty. jpm............100(0 ISO pHrriut...................... 1.9 9.4 BAcwDonate (HCOi).. 4 6 SulphAts (BOO........... 4 14 Chlorida (Cl).............. S. 16 Bilk* (BIO.)................ Ira (Tc)..................... Aluminum (At)........... Minunm (Mn>.... Calcium (Ca)............. 9 1 0 0 1 7 0.9 T Tr 6 fJO ISIS 1914 1919 1316 1466 1469 Well No Cap. ipm pH value COi HCO SO. . 1113 SO 94 0 0 36 StOiO 810.(0 FeCO Al() 11 r 10. 1944. p 1316 3700 79 300 5.9 5.4 00 44 11 9 11 19 59 3919(04!4|() Macnecium (Mi).... 1 Sodium (Na) Cate.... 1 Totd Dtaolacd......... Sottdsp)................... M Sotpcadcd Material.. 1 4 4 59 0 OBattouted (09um of the components a* determined bp an analj'ils (>) Drilled 15 ft away Ikon old well Mo 14M (ONot used to batten (*)Total la water under teat (OPiltered, total after Idtratlon (OBatbnated for water under teat Al<0 0 Ca() 4 M|(0 1 CaCO.(0 14 Tot D 8(0 91 OuaMat 10-11(0 >>4(0 >4(0 0.1 6-6(0 9-6(0 Uljtf (>)Xray eaamlnatlm) of tutpeaded oitSL Calclta, quarts and atumlnun ifllanflS HUmanlts (Ali0i.SIO) -StSi ()8um of components dstumbled by milmm water in the hot procesj would dissolve aluminum ailicate and carry aluminum to the boilers. . Thus the most important step in preparation of makeup water was re* moral of suspended material by coagu lation, preferably with a ferric-iron salt, followed by filtration at normal water temperature and at a pH of no more than 7 or 8 as an absolute mail.mum. Following clarification, we considered three procedures, (1) base exchange (2) hot-process treatment with phos phate (3) 2-etage hot-process treat ment for silica removal and phosphate softening. Objections to the base-exchange sys tem were as follows: (1) It did not fit well into the water system. Before reaching the boiler house, make-up water passes through heat-exchange equipment in plant process. Here it attains a tempera ture 160 to 170 F. This meant baseexchange softening would have to occur before, the 'makeup water was used for cooling. (2) Further changes in well-water composition might mean use of more city water or poorer quality water such as that from Well 3929 (Table I). This development would pose two problems: need for increase in base-exchange capacity; possible need for hydrogencycle exchangers in addition to sodium cycle for adjustment of alkalinity if' city-water consumption increased. (3) There would bo no provision for silica removal. While additional re moval beyond clarification might not be necessary, the specter of changing well-water composition dictated other. wise. (4) Good wells were being pumped as heavily os possible. Any increase in water use at the power plant meant greater consumption of city water. Thus the considerable quantity of water needed for rinsing base-exchange units was objectionable. Single-stage hot-process phosphate softening would reduce treated-water -hardness to a low point and minimize amount of sludge developed in boiler water. But should silica removal prove necessary, this system of* softening would provide therefor. Further, if makeup water changes made prelimi nary.hot, lime-soda softening desirable, the system would not be suitable. Tbe moat logical ayatera considered was 2-stage hot-process treatment. The first-stage hot-process unit would give some silica removal; the second stage, softening with phosphate. Should well water go bad and more city water or harder well water be needed, prelimi nary lirae-and-soda-osh softening could be accomplished in the first stage and additional phosphate softening in the second. . The equipment selected is complete, and provides for any eventuality ex cept intolerable increases in concentra tions -of ions--sulphate and chloride. equipment liutelfad. Clarifying equip ment, Fig. 2, includes a 40,000 gal tank where well water collects and mixes with enough city water to meet plant' demands. This mixture is' pumped through a meter and control valve to the 63,000-gph clarifier, where it U flocculated for removal of suspended material. Positive-displacement pumps feed ferric chloride and caustic soda, the flocculating agents, from separated .dissolving tanks to the clarifier. Ferric chloride Is automatically proportioned to the flow of water to the clarifier, and caustic-soda feed is controlled bftpHj value of clarifier effluent From ihe launders of. the cln flocculated and settled water flow a clearwoll of 4000-gal eflectivA* parity. From here it is pumped ihrjf eight 8*ff dia, vertical, pressure filfl? operating in parallel. Filtering rial is Anthrafilt. - Filters' are backwashed with filter! water and the backwash returns to]3 inlet section of ifie clarifier. No pumn are used; the pressure in tba^fillv effluent line serves to force waierSsj controlled rote of flow through^ filter beds. Filtered water goes to hent-eidstM equipment in plant process, .ffffii there it returns to a 40,000-gal tsjwSj the boiler house at 160 to 170 P/iT clarified, filtered water under, cm! of a float-operated valve maiolaig constant level In the tank. From this tank filtered w*|2 pumped, through a meter, an ainsgj ated pilot-operated inlet-water troller and a' vent condenser tgtt spray-type heater on the 63^)0Dj)Rj first-stage hot-process unit. Fig- SjjT" plant employs thoroughfare dexerj Steam vented from the deasrE heater heats makeup water to ;2f in the first-stage hot-process heater. Noncondensable geseijjj from the vent condenser. Silica' removal, for prettify water, constitutes the prime this first-siege hot-process unit. Csvg, sods solution and doloroUic-l>in*;jS}.'<5 used in the treatment are mitts separate chemical tanka. Tbefgl then pumped into the sediment# tank in proportion to the flow of-nsa Recirculation of sludge helps ttvj tation 'and silica removal. ' 90 (544) POWER $opUmbr*i| TABLE II; TREATED-WATER CONDITIONS CLARIFIER-FILTER EFFLUENT, 1947 HarJacit CaCOi. ppm Silica SlOt, ppm TvrUdity "< Net --FI*$bSlmga 4(i) MOC) reading, CaCOi, -- Sill-- -lMe it/i--- ait/n*.......... ii/4........ (l(9.. (i) lit N/19 add per 100 ml tunpta, phendphthaleln Indlcatnr. i-Staga A<> MOP). Mardnctt PharpSeta ramming, reading. CaCth, fOi. Mm ppm (0 Ml N/30 add par 100 ml amp!*, mtthjr1-orana Indicator. pH pataa 10.9 10.1 t'Seltled water from the uptake cone if.the silica-removal unit passes to .it* 63,000-gph second-stage hot-process unit for further treatment, with anhydrouj disodium phosphate. This unit ''operates flooded. Disodium phosphate solution is pumped from a 'dissolving k to the sedimentation tank, in pro portion to water flow. Phosphate-' treated water flows from tho uptake : of the phosphate softener filters mentioned above. ^Two 600-gpm pumps and ratc-of- " w controllers handle filter backwashI with softened, filtered water. Back tab returns to the top of the phos>att softener. ^Condensate returns to a collection t at a rate of about 8000 lb per hr. tp there it is pumped to the conden se compartment of the deaerating heater. Plans are progressing for reeWry'of about 35,000 lb per hr of jilant condensate. .The thoroughfare deaerator is deligned to heat 720,000 lb per hr of feedwater to a minimum of 220 F with &P*'K aieam. Oxygen concentration if-effluent is less than 0.03 ml per liter Syhe Winkler test itieg Results. Only one major difficulty has developed. JFjter quality has not been affected the operators have had `ojbo utra vigilant. The clarifier-filter installation where ijjtrinum* end silica-rich suspended -J.,8f u removed from the well-city mixture h the heart ol the treat- Floe formation has been -p' as little ferric chloride as anSi. normally 15 to 20 i|Vj' ^towgh caustio soda to hold pH ftieit" 1 We* BPPro*ching the ca* the equipment the slurry iai 1 j n 100 ^gb and then addi'Ifo haa* resulted in carryover, amount of floe to clear- cals This has happened ^tuarjy In the morning during filler backwashing and chemical-tank filling. dition is not alarming. Sludge content The alert action of the operators has of the boiler water is not high or staved off trouble and kept the filters troublesome. from being seriously fouled. They hove Table II shows typical water condi been bypassing raw water into the clari tions for a week in Dec 1947. Tur fier launders, which * reduces flow bidity, and iron eod aluminum concen through the equipment end gives the trations of the elarifier-filter effluent slurry level a chance to drop. In spite were negligible. Hardness 'as ealciuth of this bypassing procedure filtered- carbonate was 29 ppm; siliea concen quality has not suffered because tration, 8 ppm. filler ' effluent contains negligible Hardness of the first-stage effluent amounts of iron, aluminum and sus ran about one-third lower than that of pended material. the clarifier discharge. Silica concen Laboratory testa indicate that weight-, tration was reduced 50% or more oven ing ferric-hydroxide floe with a feW though pH value dropped as low as 9.3 ppm of activated silica or bentonic , toward tho end of the week. clay materially increases settling rate. Silica concentration did not change It is believed this procedure may solve in ihe second-stage effluent Hardness the difficulty described above. waa normal, between 5 and 10 ppm Silica removal and some softening hut strangely, it was higher during with hydrated dolomitlc lime and caus the first part of the week when phos tic soda at a pH close to 10.0 takes phate concentration and pH were place in the first-stage hot-process higher. Explanation is not readily ap softener. Chemical consumption varies because of irregularities in composi tion of well-city water mixture. Recent parent since filtered, second-stage ef- fluent is normally free of suspended sludge. - figures over a 15-day period are ap A word about chemical costa. Some ' proximately 40-ppm dolomitic lime and variation can be expected. During the 40-ppm caustic soda. Since silica con first 15 days of 1948 chemical costs per centration of clarified, .filtered water million lb of treated water amounted is not high, removal is limited to a few to 15.70. This is 4.8# per 1000 gal. ppm. It is arbitrarily based on holding Bailor Rnufh. The new 240,000-lb- concentration in boiler water to about per-hr steam generator has been on the 30 to 40 ppm with blowdown between line steadily since Feb 1947 and, by Feb 5 and 10% of makeup. 1948, has generated 1,512.729.000 lb Anhydrous dlsodium phosphate at 0 of ateam at an average rating of 230,000 pH in the neighborhood of 10 gives lb per hr, . peak rating 300,000 lb final hardness removal in the second- per hr. stage hot-process softener. Phosphate There have been no lube losses or. consumption changes with variations in development of analcite scale in the hardness and suspended material pres older \25,000-lb-per-hr boilers. Occa ent in the first-stage effluent. For a sional inspections have shown nothing . time it ran about 40 ppm, but tho re but a little watery sludge. cent 15-day period averaged high, al Using the internal condition of the most 55 ppm. older boilers as an index nf condi Hardness of the phosphate-softener tion! in the new unit is a unique fea effluent runs higher than expected, ture--particularly so in view of post normally between 5 and 10 ppm as experience with the older boilers and calcium carbonate. While efforts will the fact that heat release per cu ft of he continued, to reduce hardness at furnace volume is considerably greater well as phosphate consumption the eon- therein than in the new boiler. mu.at .**PUmb*r 1948 (545) 91