Document rpen3gwqYyD3qn8Xd5o4eVaOJ

DEMINERALIZATION continued TbOMtto tftmiwolljlrtg! of OdkimlMt>OQiAlnftutflt 0*1 offlbont silica W Psmvtll S-( II I fUgsiwroM S TM NaOfVcuft -- Influent AsfiwuW S-lt I'll CO* * 2 (*n ot CO* so* re pen> oi so, ACI*SO* *4 ppm 01 CCO| lCl*SO. ITepm at Co CO. a*S0,80ppn 01 CoCOj 1000 TOOO WOT OOT 9000 60OT TOOT OOTO Gan/cu fi Parmitlr SH In r>o*uro O.M. TESTS an sodium leakage effect, on final results with 2-step treatment, point out need to hold the sodium to low level Controlled Tests Point Up Difficulties in Ion Exchange anion-exchange bed before tfcou reach the flat p&rv of the tun as fA aa ilUco is concerned. In an actual plant you meet this requirement easily by length* enlng the rinse to give a low-effluent silica during the rinse recycling. But certainly for best results with 2-bed operation, sodium in the water entering the anion-exchange unit must be held as low as possible. Regenerant Caustic. Further, we found it difficult to get a good, low-siliea effluent of obout 0.01 to 0.02 ppm oi residual silica from a mixed-bed unit unless we kept regenerant dosages at about 10 lb of sodium hydroxide per cu (t of the resin, right, shove. A dos age level down around 6 lb of NaOH per cu (t produced an effluent silica ap preciably higher than 0.02 ppm. On the other hand, you con get the 0.01 to 0.02 ppm silica with a 6-lb NaOH regener ant dosage In a 2-bed demineralizer setup. During the early runs, however, while the mixed bed was still young we could readily reduce effluent silica to 0.01 ppm of SiO* or less, employing NaOH at 150 F and in sufficient dosage. But as.the bed got older It became increas ingly diffieutt to maintain this silica quality. Even acid washings foiled to restore the onion exchanger in its orig inal silica-reducing ability. Anton-Exchange Capacities. Our tests showed ua that the anion-exchanger capacity In kg per co ft is much lower in the mixed bed than in the 2-step operation. It averaged about 8 kg per cu-ft of anion removal for Permutit S-l In a mixed bed employing 10 lb of NaOH regenerant par cu ft of resin. Under the same conditions the Permutit S-l averaged 14 kg per cu ft oapacity with a 2-step system. The anion load In tho mixed bed ran higher because of the 118-ppm load from the bicarbonate alkalinity in the raw water. Between the extra anion load and the reduced exchange capacity of the Per mutit S-l In the mixed-bed design so appreciably smaller amount of water could be treated by this system per run for the same volume of anion-oxchange resin thon with a 2-step unit- Fig. 5 illustrates the considerable differences with the two methods. The results do show, however, that you eon get s de mineralized water of a silica content between 0.01- to 0.02-ppm SiO* with both the mixed- and the 2-bed methods. But it takes carefully controlled con ditions to assure these results. Sttlco and Anion Ratio. Our test water carried a high ratio of silica to total anions, about 1 to 1. 5o we have begun tests to determine the effect of this ratio. Already indications point to tho fact that the desirable, extremely low ailieo residual can be attained much easier for a 2-bed onlt handling waters with a rotfo of 1 part silica to 5 total anloaa. Tho higher this ratio gets, the more vigorous are the regenerant con ditions you have to maintain to get (he same results. From this ratio experience it would seem the mixed-bed process, with the bicarbonate* stlU in, would Improve the silica to total anion ratio, and give bet ter silica residuals. This has not proved out in practice, though. In fact, by close control of tho sodium In the In fluent to the anion exchanger it appears easier to obtain silica remora) to the 0.01-ppm level with the 2-bd titan with the mixed-bed demineralization Our tests, though, are still be ried out. Iron Effect. Dissolved iron ia fluent to the a&lon exchanger severe problems. It costs oufi exchange resin and allows lilies] uals to climb in the effluent Vi 2-ttep demineralizer the hydi lion exchanger removes i dissolved Iron In tho raw waterbed units suffer because the dlu iron comes into ready conlaet wit] anion-exchange resin. For die reason other precipitant!, a high water hordness, and measurable bldity add to the mixed-bed operiSi problems. To combat these problems, it is di able to soften at least (he regenertf and rinsing water meant for the suj exchange portion of the niatd * Frequently a Alter bed ahead oi mixed bed will prevent eventual di ging in the unit. Economic Comparison. Tsblo the typical raw waters handled by mineralisation processes. With wotcri the mixed bed has advsol Chort, at right on p 74, portrays parstlve costa except for erectioo, I and softeners for a SO-gpm afagh installation employing both mixed- 2-bed operation. When treating a low-solids water New York City's CotsVil! supplf mixed-bed equipment runs reach w then for a 2-step design. Chernies) with this raw water are not too I But when you leave the lo*4] waters (30-75 ppm) first costane*^1 leal eosts eomblne to make the w bed system relatively expensive. CNOINEEIING AND MANAGEMENT SECTION HOSPITALS Con,,. ..rfcho bonding om(d. .poco. con b. PonoU.od M -ochonlcol plan, ** >* "W-- ospitals Need Mechanical Know-How Qho-lhird the Investment in lispitol is (or heating and Jtpnlcal plant 'el costs alone vary from to $2S6 per bed a year, fijending on building deelgn .tondby boiler capacity on goes in overalls, de- 1 to meet extreme poake, 'those or average t NIIBOAARO, Howlin' Cmslunl Nwqroord, Agnow and Craig Now York and Tomato IN TliC AVEAACS HOSPITAL, COSl ol Washington, D.C Let's sec now what's heating and mechanical plant repre been going on In the hospital field and sents upwards of one-third of tho total try to draw some lessons. investment. Its operation and mainte Hospital bookkeeping usually groups nance bulk large In the yearly budget. Yet hospitala as an industry hove never analyzed their precise needs The an swer it whet you might expect--usually power, light end heat in one account. But the bulk of the expenditures in this account are for heat. Fuel, alone, amounts to about 40% ot tho total. loo much equipment for the job at hand. Tables I and II, on page 78, present a Here are some guides to good planning in this long-neglected field. For years, architects, engineers and challenging comparison. The five hospitals making up Table f spend from $74 to $}8$ a year par bed hospital building committees hsve fell the lack of simple, authoritative yard sticks to judge the requirements for power, heat, ventilation, air-condition ing, electrical equipment. Potentially the most valuable aid along these lines is the mechanical section of U.S. Public with 1950 fuel prices. Those making up Table II ruo all the way front $186 to $447 a year per bed. Major differ ence Is the tocior of Insulation. Table I entries enjoy Insulation In one-quarter to one-half the building area. The an swer seems obvious. Yet Us lesson is Health Service's publication, Functional Basis of Hospital Planning, put out by the Division of Hospital Facilities, largely overlooked. In 1947,36 hospitals in the New York area with 6906 beds spent an aggregate 1SJ ENGINEERING AND MANAGEMENT SECTION