Document p2gKQOMNyvprq7My6Bm31oG1k

Semico I Reoctioni in Demlnerolliation, Regeneration Are Straightforward, Emily Understood INO,) NOti M.I -- J [* ' 2H.CO, Co )SOorCi| N01) Co ) Z . Mg H,SO. or- 2HC> '"lllflitT Hydrogen Zrolilf Sodium - Calcium end . Mogntiium gl Sodium Ceftiumand Magnutlum Zaailta Sulfuric or cation ovchongo. tho fim o) tho two*,top do* Sodiipumon,Cdcidncigum ocid*; blcorbonotet to eorbonfc odd, wlfote* and Swcottotlon pcocoit, convert, oil Mil, to their corro- endcMhlOogrindvesl*um >0 sulfuric and hydrochloric ocid, retpectlvaly Sulfotn Of CMori^f I jrtjsa, AON H,eOt -- A HS-O.) or or } MsSlO, -- A. HCOi 1 Hfi IHCI Anion . Add In CiCAangor Ecenonger Caustic Rtqtnerottd Anion Eithanqtr . Carbonic --r or Silicic 1 Acids Anion Cictonqer holding veohodds 3Ely onion-exchange matarloll were eopable of odsorbtng only the tirongly ionised soltt, and lost tho weak ones Now strongly-baslc onlon-chonge resin* raoci on the C0 and silica, ond hence remove these troublesome weoker acids SILICA-REMOVAL demlnerolitoilon, only abAit four yeors old, was none loo soon since the latest boiler designs now on Hi ha* mode rapid progress in feedwoter treatment. Its orrivol boords caD for even higher pressures, ond purer feedeciJ Demineralization Passes in Review Here's fact-crammed summary and comparison of materials, processes to help gage the fast-moving demineralization field By S 6 APPIEBAUM, Manager, Wafer Freatmenl 0/r, Cochrone Core 4lnttmmeniotion helps determine when to regenerate and when to end rinses Regenerant measuring and Its dilution is 0 vital port of occurote eontrof over demtnerolicoiion. Here's o cose where ratio-control eflulpmenl does 0 full job Tite rapid fnoentss of deminerelfution is mainly due to the need to remove silica (or -boiler pressures abovo 500 pst. Above ibis point, silica distills over in vapor form with steam and, as the steam expends through the turbine, the silica crystallises out es a glassy deposit on turbine btades. Ultimately the tur bine must be shut down to remove this deposit by caustic scouring and, in the meantime, efficiency suffers as the silica deposit builds up. The one way to prevent this is to re duce silica in the concentrated boiler saline to very low figure*. This, in turn, calls (or extremely low silica In the feedwater to avoid excessive boiler bldwof!. . Limits of silica permissible at various boiler pressures have not been definitely determined, but el a 1951 ASME meet ing in Atlantic City, Babcock ft Wilcox es Co engineers presented their suggested silica tolerances in the concentrated boiler soltne. Table 1, p 90. From this you can calculate the limits of silica permissible in the entering makeup water to limit boiler blowdown to 5 to 10% of makeup. By dividing the silica figures in Table I by 10 to 20 respec tively, we have constructed Table II, p 90, to get these limits, Vsrlous water-treatment methods re duce hardness, alkalinity and silica In addition to demineralising, but the amount of silica these methods leave In ilio treated water is: cold-process pre cipitation, 2 to 3 ppm; hot lime teolite, 0-5 to 1 ppm; demineralising, 0.05 to 0.2 ppm. Table II indicates baiter pressures above 1000 pi! require silica in the makeup water to be between 0.25 and ENOINEERINO AND MANAGEMENT SECTION 0.5 ppm. This requirement makes * mineralising the only suitable metW to ovoid excessive holier blowdown. B* demlnerolizers are aerving for boiler operating well under 1000 psl, eru down 01 low as 600 pii, (f the witt composition ond the investment justifies it as compared to other methods. TsM* III, p 91, shows indicated methods ( water treatment for various boiler pr* sores end various raw-water compedlions. From this table you ean sea boll-, eri with operating pressures of WO 11 1000 psi have two possible choices, de mineralising and hot time zeolite. Abo** 1000-psi pressure, though, demineraliting usually proves more economical- Chemistry. Demineralizing is a 2-iW proeess. First step consists of hydrogen: cation exchange, which converts all tb* salts dissolved in the woter to their eor. . responding acids. Fig. I. The bicerbo* ; Bites are converted to carbonic acid, ^ ..which a degasifier or decarboaator libk crates; tho sulfates and chlorides proh dace ttillurie and hydrochloric acids. Any nitrates present turn up as nltrlo seidL. ;* Second step of the detnineralicing y process consists of an anion exchanger, : regenerated with caustle soda so anion \ resia is in hydroxyl form, fncldentoily, t' the first anion-exchange resins developed were weakbase realm, only capable of j adsorbing the strongly ionixed acids, : such as sulfuric and hydrochloric acid, \ Fit- 2. Bot, jbese weakly basic ex- f thtngtis could not remove weakly dls* | tecUted acids, like carbonic or ailicio irfds and, therefore, were not valuable for boiler-feed application. Fig. 3 shows "O'* the new sirongly-basic anion ex changers read to remove CO* end silica, and hence make demineralization im portant in tho boiler field. Cation Materials, For the first step, the hydrogen-cation exchange, there are today three successful materials: I. SuUonoled cooI, oldest of the hy drogen zeolites dating back to 1935, has the advantage o( easy regeneration. It requires only a single strength of 2% sulfuric acid, introduced rapidly to avoid calcium sulfate crystallization on the ezchange material. But its capacity is lower than the other available mate rials and, therefore, requires larger con tainers. Its cost, though, is lower and this frequently compensates for the larger sholl. It is sensitive to high amounts of chlorino, say over 0.3 ppm, which may attack the material. But un der the proper conditions and far waters EMOINEECINO AND MANAGEMENT SECTION relatively high in percent of alkalinity present, it is a very good material to use. 2. Medium-capacity resin, such as that made by Rohm ft Haas, as their IR-112. and Natco-Dow, as their Nslcite MCR, is relatively new, only about a year old. Like sulfonated coal it requires only 0 single strength of acid in regeneration. It cornea as styrene beads but somewhat softer than those of the high-capacity resins to be described presently. As a result you may experience higher pres sure losses. What's more, the bead par ticle swells during the hydrogen ex change run and contracts during the regeneration. So you have to allow for these volume changes. Further field 'experience will eventually determine whether these volume changes affect the long-term stability of the material, fContinued on page 90) 19