ebook img

Passive Aerobic Treatment of Net-Alkaline, Iron-Laden Drainage from a Flooded Underground PDF

29 Pages·2017·1.42 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Passive Aerobic Treatment of Net-Alkaline, Iron-Laden Drainage from a Flooded Underground

UUnniivveerrssiittyy ooff NNeebbrraasskkaa -- LLiinnccoollnn DDiiggiittaallCCoommmmoonnss@@UUnniivveerrssiittyy ooff NNeebbrraasskkaa -- LLiinnccoollnn USGS Staff -- Published Research US Geological Survey 2007 PPaassssiivvee AAeerroobbiicc TTrreeaattmmeenntt ooff NNeett--AAllkkaalliinnee,, IIrroonn--LLaaddeenn DDrraaiinnaaggee ffrroomm aa FFllooooddeedd UUnnddeerrggrroouunndd AAnntthhrraacciittee MMiinnee,, PPeennnnssyyllvvaanniiaa,, UUSSAA Charles A. Cravotta III US Geological Survey, PA Water Science Center, 215 Limekiln Rd, New Cumberland, PA, USA, [email protected] Follow this and additional works at: https://digitalcommons.unl.edu/usgsstaffpub Part of the Earth Sciences Commons Cravotta III, Charles A., "Passive Aerobic Treatment of Net-Alkaline, Iron-Laden Drainage from a Flooded Underground Anthracite Mine, Pennsylvania, USA" (2007). USGS Staff -- Published Research. 275. https://digitalcommons.unl.edu/usgsstaffpub/275 This Article is brought to you for free and open access by the US Geological Survey at DigitalCommons@University of Nebraska - Lincoln. It has been accepted for inclusion in USGS Staff -- Published Research by an authorized administrator of DigitalCommons@University of Nebraska - Lincoln. MineWaterEnviron(2007)26:128–149 DOI10.1007/s10230-007-0002-8 TECHNICAL ARTICLE Passive aerobic treatment of net-alkaline, iron-laden drainage from a flooded underground anthracite mine, Pennsylvania, USA Charles A. Cravotta III Received:3April2007/Accepted:1May2007/Publishedonline:18August2007 (cid:1)Springer-Verlag2007 Abstract Thisreportevaluatestheresultsofacontinuous June 2005. During the first 12 months of operation, esti- 4.5-daylaboratoryaerationexperimentandthefirstyearof mateddetentiontimesinthetreatmentsystemrangedfrom passive, aerobic treatment of abandoned mine drainage 9 to 38 h. However, in contrast with 80–100% removal (AMD)fromatypicalfloodedundergroundanthracitemine of Fe2+ over similar elapsed times during the laboratory in eastern Pennsylvania, USA. During 1991–2006, the aeration experiment, the treatment system typically AMD source, locally known as the Otto Discharge, had removed less than 35% of the influent Fe2+. Although flowsfrom20to270 L/s(median92 L/s)andwaterquality concentrations of dissolved CO decreased progressively 2 that was consistently suboxic (median 0.9 mg/L O ) and within the treatment system, the PCO values for treated 2 2 circumneutral(pH&6.0;netalkalinity>10)withmoderate effluent remained elevated (10(cid:1)2.4 to 10(cid:1)1.7atm). The ele- concentrations of dissolved iron and manganese and low vatedPCO maintainedthepHwithinthesystematvalues 2 concentrationsofdissolvedaluminum(medians of11,2.2, less than 7 and hence slowed the rate of Fe2+ oxidation and <0.2 mg/L, respectively). In 2001, the laboratory aer- compared to the aeration experiment. Kinetic models of ation experiment demonstrated rapid oxidation of ferrous Fe2+oxidationthatconsidereffectsofpHanddissolvedO 2 iron(Fe2+)withoutsupplementalalkalinity;theinitialFe2+ were incorporated in the geochemical computer program concentrationof16.4 mg/Ldecreasedtolessthan0.5 mg/L PHREEQC to evaluate the effects of detention time, pH, within 24 h; pH values increased rapidly from 5.8 to 7.2, and other variables on Fe2+ oxidation and removal rates. ultimately attaining a steady-state value of 7.5. The These models andthe laboratory aeration experiment indi- increasedpHcoincidedwitharapiddecreaseinthepartial catethatperformanceofthisandotheraerobicwetlandsfor pressure of carbon dioxide (PCO ) from aninitial value of treatment of net-alkaline AMD could be improved by 2 10(cid:1)1.1atmtoasteady-statevalueof10(cid:1)3.1atm.Fromthese aggressive, continuous aeration in the initial stage to results, a staged aerobic treatment system was conceptual- decreasePCO ,increasepH,andaccelerateFe2+oxidation. 2 izedconsistingofa2 mdeeppondwithinnovativeaeration and recirculation to promote rapid oxidation of Fe2+, two Keywords Aerobic wetlands (cid:2) Iron oxidation (cid:2) 0.3 mdeepwetlandstofacilitateironsolidsremoval,anda Carbon dioxide exsolution (cid:2) Kinetics modeling supplementaloxiclimestonedrainfordissolvedmanganese and trace-metal removal. The system was constructed, but without the aeration mechanism, and began operation in Introduction Problem Electronicsupplementarymaterial Theonlineversionofthis article(doi:10.1007/s10230-007-0002-8)containssupplementary material,whichisavailabletoauthorizedusers. Drainage from abandoned coalmines degrades more than 8,000 km of streams in the northern Appalachian region, C.A.CravottaIII(&) USA;manyofthesestreamsarefishlessorcontainfewfish USGeologicalSurvey,PAWaterScienceCenter, (U.S. Environmental Protection Agency 2006; Herlihy 215LimekilnRd,NewCumberland,PA,USA et al. 1990). Although acidic pH (<4.5) and elevated e-mail:[email protected] 123 MineWaterEnviron(2007)26:128–149 129 concentrations of dissolved sulfate, iron, and other metals from typical Bituminous Coalfield AMD. Studies at wet- are common characteristics of abandoned mine drainage lands receiving AMD have indicated that the prescribed (AMD), a substantial percentage of the AMD in the iron-removal rate of 20 g/m2/day can underestimate the Northern Appalachian Region is iron-laden but not acidic areaneededforeffectivetreatment,particularlyaspHand/ (Cravottaetal.1999;HymanandWatzlaf1997;Kirbyand or concentrations of iron decrease (Kirby et al. 1999; Cravotta2005a;RoseandCravotta1998).Forexample,in Tarutis et al. 1999; Watzlaf et al. 2004). Generally, flow 1999, 30% of 140 sampled AMD sources in the Bitumi- rates are higher and concentrations of acidity, sulfate, and nous and Anthracite Coalfields of Pennsylvania had iron are lower for similar pH in the Anthracite Coalfield concentrations of dissolved iron greater than the 7 mg/L compared to the Bituminous Coalfield of Pennsylvania criteria for effluent from active mines in Pennsylvania (Cravotta2007). (Commonwealth of Pennsylvania 2002) and had net-alka- The oxidation rate of iron in aqueous systems is com- linewaterquality,withnear-neutralpH(6.0–7.5)(Cravotta plex and tends to increase with water temperature, pH, 2007; Kirby and Cravotta 2005b). Despite their net-alka- concentration of dissolved oxygen, concentrations of dis- line water quality, the dissolved iron tends to precipitate solvedandsuspendediron,andbacterialactivity(Dempsey downstream from the AMD sources, forming ochreous et al. 2001; Kirby et al. 1999; Stumm and Lee 1961; encrustations that degrade the aquatic habitat (Earle and Stumm and Morgan 1996; Watzlaf et al. 2001 2004; Wil- Callaghan 1998; Williams et al. 2002; Winland et al. liamson et al. 1992). In the presence of dissolved oxygen 1991). Various alternatives for treatment to remove the (O ), concentrations of dissolved iron tend to decrease 2 dissolved iron and associated metals from AMD before it because of oxidation and hydrolysis reactions that lead to discharges to streams could be appropriate depending on the formation of various ferric-hydroxide and hydroxysul- thevolumeoftheminedischarge,itsalkalinityandacidity fate solids, hereinafter denoted Fe(OH) : 3 balance, and the available resources for construction and maintenance of a treatment system (Hedin et al. 1994; Fe2þþ0:25O2þ2:5H2O!FeðOHÞ3ðsÞþ2Hþ ð1Þ Skousen et al. 1998). Information on the hydrologic vari- Generally, the ferrous iron (Fe2+) oxidation step is rate- ability and the site characteristics is needed by water- limiting; ferric iron (Fe3+) hydrolysis and precipitation of resource managers to identify effective remedial strategies Fe(OH) arenearlyinstantaneous(SingerandStumm1970; for the AMD. 3 Stumm and Morgan 1996). Despite initial pH values that can be near neutral, some AMD containing dissolved iron canhaveacidicpH(<4.5)aftercompleteoxidationbecause Background of the release of protons (H+) associated with iron hydrolysis. As equation 1 proceeds to completion, the pH Current guidelines for passive treatment of coalmine willdecreaseunlessbufferedbyexcessalkalinity,indicated drainage (Skousen et al. 1998; Watzlaf et al. 2004) were by positive values for net alkalinity or negative values for largely developed by the former U.S. Bureau of Mines hotperoxide acidity (Kirbyand Cravotta 2005a,b). (Hedin et al. 1994) on the basis of systems that were con- The pH, alkalinity, and associated properties of AMD structedfortreatmentofAMDintheBituminousCoalfield also can change as a water sample equilibrates to atmo- of Pennsylvania and West Virginia. These guidelines spheric conditions because of the exsolution of dissolved indicate different treatment strategies for net-alkaline and carbon dioxide (CO ). Ground water and coalmine drain- net-acidic water quality. For AMD that has net acidity 2 age commonly contain elevated concentrations of (alkalinity < acidity; hot peroxide acidity > 0), treatment dissolvedCO ,resultinginpartialpressureofCO (PCO ) systemsthatpromotealkalinityproductionareappropriate. 2 2 2 of10(cid:1)1.5to10(cid:1)0.5atminthevadosezone(Langmuir1997; The alkalinity can be produced by limestone dissolution Rose and Cravotta 1998). When exposed at the surface, and, in some cases, sulfate reduction (e.g., Hedin et al. excess CO will exsolve until dissolved CO concentra- 1994;Langmuir1997).Net-alkalineAMDdoesnotrequire 2 2 tions equilibrate with atmospheric PCO of approximately supplemental alkalinity and can be routed directly to an 2 10(cid:1)3.5 atm. To explain effects on pH and alkalinity, the oxidation pond(s) and/or aerobic wetland(s) where dis- CO exsolution process may be described as: solved iron is oxidized and precipitated from solution. 2 Typically, to determine the areal extent of an aerobic sys- HCO(cid:1) $CO ðgÞþOH(cid:1); ð2Þ tem, the average iron-loading rate of the AMD source 3 2 expressed in grams per day (g/day) is divided by an where equation 2 goes to the right. The pH will increase assumed iron-removal rate of 20 g/m2/day, based on the as CO exsolves. Exsolution of CO can be promoted 2 2 work of Hedin et al. (1994). However, this areal-removal by aeration within a stream or aggressive air sparging ratemaynotbeapplicableforAMDthatdiffersincharacter in a treatment system (Jageman et al. 1988). Unless 123 130 MineWaterEnviron(2007)26:128–149 accompanied by the precipitation of Fe(OH) , CaCO , or thebasisforthedesignofaerobicwetlandsfortreatmentof 3 3 other solids, the alkalinitywill notbeaffected by equation the Otto Discharge. The performance of the treatment 2 because bicarbonate (HCO(cid:1)) and hydroxyl (OH(cid:1)) ions system is evaluated on the basis of water-quality data 3 have equivalent capacity to neutralize acid; that is, they collected monthly during June 2005 to June 2006 at the have equivalent alkalinity. On the other hand, if Fe(OH) inflow, outflow, and intermediate points within the treat- 3 precipitates in accordance with Eq. 1, alkalinity will ment systemand onthe receiving stream above andbelow decrease and the pH also could decrease, but not neces- the discharge. Lastly, geochemical models of iron oxida- sarily. For solutions containing dissolved iron and excess tion are used to evaluate detention time and chemical CO , the pH could increase or decrease depending on the factors affecting iron oxidation and removal. 2 concentrationsoftheseconstituentsandtherelativeratesof reactions (1) and (2). Wetlands designed for removal of iron may be inade- Description of Study Area quate for the removal of dissolved manganese and various trace metals. Alkaline pH values may be necessary for The Otto Colliery Discharge (Otto Discharge; lat. attenuation of these metals. At neutral to alkaline pH, iron 40(cid:2)3905800 N, long. 76(cid:2)1901400 W) near Branchdale, and manganese oxides can be effective sorbents of dis- SchuylkillCounty,isoneofthelargestsourcesofAMDin solved manganese, copper, cobalt, nickel, zinc, cadmium, theSchuylkillRiverBasinineasternPennsylvania(Fig. 1). and other potentially toxic trace metals (Kooner 1993; The upper Schuylkill River Basin originates within the McKenzie 1980). Such metals have been identified with Southern Anthracite Coalfield of the Appalachian Moun- ochreous precipitates at AMD sites (Kairies et al. 2005; tain Section of the Ridge and Valley Physiographic Webster et al. 1998; Williams et al. 2002; Winland et al. Province, where it is underlain by strongly folded and 1991). Adsorption of the cations by the hydrous iron and faulted sandstone, shale, siltstone, conglomerate, and coal manganeseoxidesgenerallyisinterpretedtobetheprimary of the Llewellyn and Pottsville Formations (Wood et al. mechanismforremovalofthemetals.Forexample,cobalt, 1986; Eggleston et al. 1999; Way 1999). Since the late nickel, and zinc in AMD were attenuated under oxidizing 1800s, a total of 38 mapped coalbeds with average thick- conditions within a buried limestone bed, but only after nesses from 0.3 to 2.5 m have been mined to depths ironandmanganeseoxidesprecipitatedneartheinflowhad exceeding 1,000 m in the Southern Anthracite Coalfield migrated downgradient where the pH was near neutral (Wood et al. 1986). Although several surface and under- (Cravotta and Trahan 1999; Cravotta and Watzlaf 2002). ground anthracite mines presently (2007) are active, most Althoughelevatedtemperaturescanproducefasterrates minesintheupperSchuylkillRiverBasinwereabandoned of iron oxidation and more effective treatment in AMD before 1960 and are flooded. Fresh surface and ground wetlands (Dempsey et al. 2001; Kirby et al. 1999), the water that enters the mines acquires acidity, sulfate, iron, prolongedexposureofAMDtoambientairtemperaturesor and other metals and eventually discharges as AMD from sunshine can produce temperature extremes that are not mine shafts, tunnels, and topographically low points suitable for fish and other aquatic organisms. Directing (Growitzetal.1985;RobertKimballandAssociates2000; effluentfromashallowwetlandthroughatrenchfilledwith Pennsylvania Department of Environmental Protection limestone or other rock fragments can moderate the water 2003; Wood 1996). The metal-laden AMD contributes temperature. Ideally, the treated AMD effluent quality and substantially to base flow but degrades stream-water temperature would sustain fish and other aquatic life. quality and aquatic habitat of the Schuylkill River and its upper tributaries. Consequently, the West Branch and upper Schuylkill River are on the U.S. Environmental Purpose and Scope ProtectionAgency(USEPA)303(d)listofimpairedwaters in Pennsylvania (Pennsylvania Department of Environ- This report presents information that could be useful to mental Protection 2003, 2004). water-resource managers to assess remedial strategies for The Otto Mine operated from the early 1800s to 1934 net-alkaline, iron-laden AMD. The report describes field, and encompassed an area of approximately 20 km2 (Guy laboratory, and computational methods and water-quality Mace, White Pine Coal Company, oral communication, data collected by the U.S. Geological Survey (USGS) 2006).Duringthelifeofthemine,thousandsofmetrictons during 2001–2006 for the evaluation of passive-treatment of anthracite were extracted from 13 named coal beds strategiesandperformanceresultsatafloodedunderground (RabbitHolethroughBuckMountain)toamaximumdepth anthracite mine in eastern Pennsylvania that is locally of60 mbelowsealevel(Egglestonetal.1999;GuyMace, known as the Otto Discharge. A laboratory aeration WhitePineCoalCo.,oralcommunication,2006).Afterthe experiment conducted in December 2001 is presented as Otto Mine closed, the underground voids flooded and the 123 MineWaterEnviron(2007)26:128–149 131 Fig.1 Maps showing location of the Otto Discharge. a Bituminous County,Pa.CoalfieldboundariesbasedondistributionofPennsylva- and Anthracite Coalfields of Pennsylvania; b Abandoned mine nianbedrock(Bergetal.1980).MapofupperSchuylkillRiverBasin drainage (AMD) in the upper Schuylkill River Basin, Schuylkill adaptedfromL.RobertKimballandAssociates,Inc.(2000) present discharge developed from a production tunnel and Datumof1929(NGVD29)andflowsdowna470 mlong, associatedshaft.TheuntreatedAMDexitsthetunnelatan rock-inlaid ditch to elevation 238 m above NGVD 29 at elevation of 250 m above the National Geodetic Vertical Muddy Branch (Fig. 1). At their confluence, the Otto 123 132 MineWaterEnviron(2007)26:128–149 DischargecontributesthemajorityofstreamflowinMuddy coagulation and settling, plus oxidation rates could be Branch, which flows eastward to West Creek and eventu- increased by sorption of Fe2+ and catalysis on Fe(OH) 3 ally the West Branch of the Schuylkill River. Below their surfacesatnear-neutralpH.Theabioticoxidationofsorbed confluence, the water and streambed of Muddy Branch are Fe2+ (heterogeneous oxidation) generally is faster than the a rusty orange-brown, owing to loading of iron and other oxidation of dissolved Fe2+ (homogeneous oxidation) metals from the Otto Discharge. Despite the metals load- (Dempsey et al. 2001; Dietz and Dempsey 2002; Kirby ing, the Pennsylvania Fish Commission (Chikotas and et al. 1999; Stumm and Sulzberger 1992; Tamura et al. Kaufman2002)documentedhighdensitiesofnativebrook 1976).Enhancingironremovalbyheterogeneousoxidation trout (Salvelinus fontinalis) in Muddy Branch downstream could improve treatment efficiency and produce a denser from the Otto Discharge. The brook trout benefit from sludge. sustainedinflows ofcold,near-neutralwaterfromtheOtto Landowner consent was obtained, detailed engineering Discharge; however, populations of aquatic insects and was completed, construction permits were approved, and, other fish species tend to be limited because of iron-oxide in June 2005, construction of the treatment system was depositsinthestreambed(e.g.,EarleandCallaghan1998). completed by private-sector companies on behalf of the Data collected by the USGS from 1991 to 2006 (this Schuylkill Headwaters Association, Inc. The ‘‘as-built’’ paper; Williams et al. 2002; Wood 1996) indicate that the system incorporated most basic features of the conceptual OttoDischargerangeswidelyinflow(20–270 L/s;median design. However, to reduce construction costs, two novel 92 L/s) but is consistently net alkaline (pH = 5.7–6.1; provisions—the hydraulic aeration and recirculation in the alkalinity = 34–70 mg/L CaCO ), anoxic to suboxic upperoxidationpondandtheperforatedflushpipinginthe 3 (<0.1–2.2 mg/L O ), and contaminated with sulfate loweroxiclimestonedrain—wereeliminated.Asshownin 2 (SO = 190–370 mg/L), iron (Fe = 3–16 mg/L), manga- Fig. 2, the AMD from the Otto Discharge was diverted 4 nese (Mn = 2–3 mg/L), aluminum (Al = 0.1–1.3 mg/L), fromtherock-linedditchatabout235 mdownstreamfrom and other dissolved metals. On the basis of its net-alkaline thetunnelandspilledsequentiallyintotheoxidationpond, waterqualityandflowcharacteristics,theeffluentfromthe the two shallow wetlands, and finally the oxic limestone Otto Discharge could be treated with a passive aerobic drain. Rectangular weirs with adjustable freeboard were wetland system for removal of iron and other metals (e.g., installed between each of the treatment cells to control Hedin et al. 1994). water levels and facilitate flow measurements. At the sec- ondwetland,anoverflowweirwasinstalledasasecondary outlet in the event the oxic limestone drain would not be Treatment System Design and Construction able to transmit the entire flow. A passive aerobic treatment system was conceptualized to facilitatetheremovalofiron,manganese,andothermetals Materials and Methods from the Otto Discharge while maintaining the sustained flows of oxygenated, cold water required by native brook Water-Quality Sampling and Analysis trout in downstream reaches of Muddy Branch. The con- ceptual treatment system would divert the existing Data on the flow rate and inorganic chemistry of the Otto discharge with a low-head dam; promote aeration with an Discharge and Muddy Branch above and below the Otto in-linehydraulicdevice;detaintheaeratedwaterwithinan Discharge were collected by the USGS over a range of oxidation pond; promote deposition of iron-hydroxide hydrologicconditionsduring2001to2006.Oncetreatment precipitates that form as a consequence of aeration within began, additional water-quality data were collected at the the oxidation pond and two constructed, aerobic wetland outletsofeachofthetreatmentcells.Aftertreatmentwasin cells; and remove manganese and trace metals in a flush- placefromlateJune2005toearlyJune2006,allthewater able, oxic limestone drain that would add alkalinity and fromtheOttoDischargewasdivertedthroughthetreatment mitigate temperature fluctuations. The proposed in-line system. Sample sites along the rock-lined diversion ditch hydraulicaerationwouldtakeadvantageofgravityanduse from the Otto tunnel were at the inflow (Otto 235-m) and venturinozzlesnearthepipeoutlettodrawairintothepipe below the outflow (Otto 470-m) of the treatment system (e.g., Ackman and Kleinmann 1985), hence, supersatura- and,hence,usefultoindicatethe‘‘untreated’’and‘‘treated’’ tingtheinfluentwithatmosphericO beforedischargingto waterquality,respectively.Monitoringsitedescriptionsare 2 the first treatment cell. The aerated influent would be reportedin Table 1. injected near the bottom of the oxidation pond to promote TheflowrateofMuddyBranchaboveandbelowtheOtto the circulation of the water and Fe(OH) particles. The Dischargewasmeasuredbyusingawadingrodandpygmy 3 circulation of Fe(OH) particles could facilitate particle currentmeter(Rantzetal.1982a).Afterthetreatmentsystem 3 123 MineWaterEnviron(2007)26:128–149 133 Fig.2 Schematic plan of passive, aerobic treatment system con- stages;ababove,blbelow.Dataontheareaandvolumecapacityare structedin2005attheOttoDischarge,Branchdale,Pa.Labelledpoint for the design dimensions, slopes, and surface-water elevations symbols (e.g., Otto-235m) indicate water-quality sample sites; (ChristineHaldemanandClaytonBubeck,RettewAssociates,written arrows indicate general flow direction through sequential treatment communication,2006) wasinstalled,weirsattheoutletofeachtreatmentcellwere were assumed constant because only minor changes in usedtoestimatetheflowratethroughoutthetreatmentsys- water depth were recorded at the weirs between cells tem. The flow across the weir notch was estimated on the compared to the total depths of the cells. basisofastandardequationforrectangularnotchgeometry At each sample site, temperature, pH, specific conduc- (Rantzetal.1982b)adaptedformetricunits: tance (SC), dissolved oxygen (DO), and redox potential Flowrate(cid:1)Ls(cid:1)1(cid:2)¼ð1:839(cid:2)widthÞ(cid:2)ðdepthÞ1:5 ð3Þ (Eh) were measured by using a multiparameter, sub- mersible sonde. The sonde was calibrated daily in where the width is 1.52 m and the water depth over the accordancewithstandardmethods(U.S.GeologicalSurvey weir is in centimeters. To estimate detention time within 1997 to present). Field pH and Eh were determined by each treatment cell, the volume of water contained within using a combination Pt and Ag/AgCl electrode with a pH thecell(Fig. 2)wasdividedbytheflowrate.Thevolumes sensor. The electrode was calibrated in pH 4.0 and 7.0 Table1 Descriptionsofwater-qualitymonitoringsitesonMuddyBranchandtheOttoDischargetreatmentsystematBranchdale,Pa USGSStationID Sitenameanddescription Longitudea Latitude Elevationbm 403958076191401 OttoMineTunnelatBranchdale (cid:1)76.32022 40.66620 250 0146784348 Otto235m‘‘Inflow’’ (cid:1)76.31939 40.66758 244 404001076191301 OttoTreatmentCell1‘‘OxidationPond’’ (cid:1)76.32033 40.66705 244 404006076191001 OttoTreatmentCell2‘‘Wetland1’’ (cid:1)76.31927 40.66803 244 404005076190901 OttoTreatmentCell3‘‘Wetland2’’ (cid:1)76.31956 40.66842 244 404008076190601 OttoTreatmentOxicLimestoneDrain (cid:1)76.31842 40.66875 244 0146784350 Otto470m‘‘Outflow’’ (cid:1)76.31828 40.66870 238 0146784338 MuddyBranch0.1kmaboveOtto (cid:1)76.31861 40.66889 244 0146784354 MuddyBranch1.0kmbelowOtto(PF101),SteinsMillRd (cid:1)76.30666 40.66750 232 0146784358 MuddyBranch1.8kmbelowOtto(PF102),BlackDiamondRd (cid:1)76.30278 40.66472 230 ApproximatesitelocationsshowninFig.2 a HorizontalcoordinateinformationisreferencedtotheNorthAmericanDatumof1983(NAD83) b ElevationsreferencedtotheNationalGeodeticVerticalDatumof1929(NGVD29) 123 134 MineWaterEnviron(2007)26:128–149 buffer solutions and in ZoBell’s solution (Wood 1976, pp. described by Barbour et al. (1999) and Meador et al. 18–22; U.S. Geological Survey 1997 to present). Values (1993). Although benthic-macroinvertebrate samples also for Eh were corrected to 25(cid:2)C relative to the standard were collected during the annual surveys, sample pro- hydrogen electrode in accordance with methods of Nord- cessingandtaxonomicidentificationwerenotcompletedat strom (1977). Unfiltered and filtered (0.45 lm pore size) the time of this report (2007). samplesofwaterwereprocessedinthefield,transferredto polyethylene bottles, preserved as appropriate, and trans- ported on ice to the laboratory. Laboratory Aeration Experiment Thealkalinityand‘‘hotperoxide’’acidity(hotacidity)of theunfilteredwatersamplesweretitratedusingsulfuricacid AlaboratoryaerationexperimentwithAMDfromtheOtto (H SO )and/orsodiumhydroxide(NaOH)tofixedendpoint 2 4 Discharge was conducted in December 2001 to determine pH of 4.5 and 8.3, respectively (American Public Health the potential rate ofironremoval andtoverify that thepH Association 1998a, 1998b). Typically, alkalinities were could be sustained near neutrality without supplemental measured within 24 h of sampling, whereas acidities were alkalinity.OnthemorningofDecember3,2001,rawwater measured several days later at the USGS Water Science and corresponding field measurements of water quality CenterlaboratoryinNewCumberland,Pennsylvania.Con- were collected from the discharge as it exited the Otto centrationsofmajoranions(SO ,Cl)infiltered,unpreserved 4 Minetunnel.Atotalof30 Lofrawwaterwascollectedin subsampleswereanalyzedbyionchromatography(IC),and a66 Lcapacityplasticcoolerandimmediatelytransported concentrationsofmajorcations(Ca,Mg,Na,K)andselected to the author’s residence. An aquarium pump with porous trace metals(Fe, Mn,Al,Ni,Zn)inunfilteredandfiltered, stone diffuser set at the bottom of the cooler delivered a acidifiedsubsampleswereanalyzedbyinductivelycoupled continuousair-flowrateof1.5 L/min;therisingairbubbles plasma optical emission spectrometry (ICP) at the Actlabs promoted aeration and circulation of the water and sus- laboratory in Toronto, Ontario, or the USGS Mineral pended fine particles. The ambient temperature of the Resources Laboratory (MRL) in Denver, Colorado (Crock initial sample, 12.3(cid:2)C, was approximately maintained etal.1999;FishmanandFriedman1989). during the aeration experiment (10.5–13.7(cid:2)C). A multipa- To supplement data on hot acidity, the net acidity was rametersondewassubmergedatthebottomofthecoolerto computed as mg/L of CaCO , considering positive acidity 3 record temperature, pH, SC, DO, Eh, and turbidity. Serial contributions from pH and concentrations of dissolved sampling was conducted to evaluate changes in alkalinity, iron, manganese, and aluminum in mg/L (C , C , and Fe Mn acidity, and concentrations of metals. At fixed time inter- C , respectively), and negative contributions from alka- Al vals,a500 mLpolyethylenebottlewasimmersedopposite linity as mg/L of CaCO : 3 the aerator and its contents were processed for chemical analysis. Alkalinity and ‘‘cold peroxide’’ acidity (cold NetAcidity acidity)inunfilteredsamplesweremeasuredinUSGSNew ¼50(cid:3)10ð3(cid:1)pHÞþ2(cid:2)C =55:85þ2(cid:2)C =54:94þ3(cid:2)C =26:98(cid:4) Cumberlandoffice.Thecoldacidityincludescontributions Fe Mn Al from the dissolved CO plus dissolved iron and aluminum 2 (cid:1)Alkalinity ð4Þ butmayexcludecontributionsfrommanganese(Kirbyand Cravotta 2005a, b); hot acidity was not measured because Kirby and Cravotta (2005a, b) showed that net acidity the computed net acidity (Eq. 4) was assumed to provide computedwithEq.4iscomparableinvaluetothestandard equivalent information. Concentrations of dissolved methodhotaciditywheretheH SO addedtothesampleis 2 4 (0.22 lm filter) and total metals in acidified samples were subtractedfromtheNaOH added(AmericanPublicHealth measured by ICP at the U.S. Department of Energy labo- Association 1998a). ratory in Pittsburgh, Pennsylvania. Annual aquatic ecological surveys were conducted in October 2002, 2003, 2004, and 2005 on Muddy Branch at one site immediately upstream and two sites downstream from the confluence with the Otto Discharge. The two Geochemical Models of Iron-Oxidation Rates downstream stations (PF 101 and PF 102 in Table 1) were surveyed previously by the Pennsylvania Fish and Boat Variouskineticoxidationmodelsthatconsideredeffectsof Commission (Chikotas and Kaufman 2002). Fish were detention time (travel time), pH, O , and Fe2+ concentra- 2 collected by electrofishing over a 150-m reach consisting tion were used to evaluate the data for the laboratory of mixed riffle, run, and pool habitats at each stream site, aerationexperimentandtheOttotreatmentsystem.Stumm held for measurement and identification, checked for and Lee (1961) proposed a widely cited rate law for the anomalies, and then released in accordance with methods abiotic oxidation of Fe2+ under laboratory conditions: 123 MineWaterEnviron(2007)26:128–149 135 d(cid:5)Fe2þ(cid:6)=dt¼(cid:1)k(cid:5)Fe2þ(cid:6)½OH(cid:1)(cid:3)2PO ; ð5Þ thermodynamicdatabase usedbyWATEQ4Fwasused for 2 the PHREEQCsimulations. where [Fe2+]denotes concentration (inmol/L)andkis the rate constant, 1.33 · 1012L2/(mol2 atm s). Equation 5 indicates the abiotic oxidation of Fe2+ is strongly Results: Laboratory and Field Observations dependent on pH; the rate increases by a factor of 100 witheachunitincreaseinpH.Incontrast,Williamsonetal. Laboratory Aeration Experiment (1992) reported that the biological oxidation rate for Fe2+ in surface waters receiving AMD is independent of pH Initially, the Otto Discharge sampled for the aeration overtherange2.5–6.0andfasterthantheabioticoxidation experiment was colorless and anoxic, with pH of 5.8 and rate as indicated by: dissolved iron concentration of 16.4 mg/L. Aggressive aeration promoted the rapid oxidation of dissolved Fe2+; d(cid:5)Fe2þ(cid:6)=dt¼(cid:1)k(cid:5)Fe2þ(cid:6); ð6Þ freshly precipitated, suspended Fe(OH) particles tempo- 3 rarily changed the water to a deep orange-brown color. In where the rate constant is 10(cid:1)3.61 s(cid:1)1. less than 24 h, the initial Fe2+ concentration of 16.4 mg/L Forthisstudy,thetwooxidationmodelswereevaluated decreasedtolessthan0.5 mg/L(Fig. 3a).Nevertheless,the in a spreadsheet by solving the integrated form of Eqs. 5 concentrations of total iron declined slowly because of and 6: slow settling of Fe(OH) particles. The abundance of sus- 3 pended Fe(OH) particles was indicated by the turbidity, ½Fe(cid:3) ¼½Fe(cid:3) expf(cid:1)k0(cid:2)tg ð7Þ 3 t t¼0 which increased from zero to a maximum value of 230 wherek0 fortheintegratedformofEq.5includesspecified nephelometric turbidity units (NTU) after 12 h of aeration constant values of pH and PO . For the specified initial and then gradually declined as the Fe(OH) settled from 2 3 concentration of Fe2+, the spreadsheet models computed suspension(Fig. 3a).After4.5 daysofcontinuousaeration, the concentration of residual Fe2+ as a function of elapsed most Fe(OH) particles had settled to the bottom of the 3 time,t.Thespreadsheetmodels,equivalenttothemodeling cooler and the supernatant was again colorless. approach used in the ‘‘AMDTreat’’ model for mine-drain- Once aggressive aeration began (1 h after collection), age treatment (U.S. Office of Surface Mining Reclamation theDOincreasedtosaturation(Fig. 3c).Althoughthecold andEnforcement2002),didnotconsidereffectsontherate temperature of the discharge water was maintained during constant owing to variations in the iron concentration, pH, the aeration experiment (10.5–13.7(cid:2)C), aeration promoted or temperature as oxidation progresses. evaporation and the exsolution of dissolved CO . The SC, 2 To evaluate interactions among the dissolved iron con- which indicates dissolved solute concentrations, declined centration and oxidation state, pH, alkalinity, and other during the first 12 h of aeration due to the precipitation of variables,thetwokineticmodelsofFe2+oxidation(Eqs.5, Fe(OH) and the conversion of HCO(cid:1) to CO : 3 3 2 6) also were used in the geochemical computer program PHREEQC (Parkhurst and Appelo 1999). This model, Fe2þþ0:25O þ2HCO(cid:1)þ0:5H O¼FeðOHÞ þ2CO ðgÞ 2 3 2 3 2 given in Appendix A (in Electronic supplementary ð8Þ material), was adapted from ‘‘Example 9—Kinetically controlled oxidation of ferrous iron’’ provided by the but thereafter increased in accordance with the sulfate PHREEQC authors. Initial models for this study assumed concentrations (Fig. 3b). Because of evaporation, concen- constant, equilibrium values of PO and/or PCO and lim- trations of sulfate increased by about 5% during the 2 2 ited concentrations of dissolved Fe3+ to equilibrium with experiment. amorphous Fe(OH) . Subsequent models simulated atmo- Despitethenearlycompleteoxidationandhydrolysisof 3 sphericdisequilibriumwithrespecttoCO byincorporating iron, net-alkaline character was maintained during the 2 atermfortheobservedasymptoticdecreaseinPCO during aeration experiment. Initially, the Otto Discharge had pH 2 the aeration experiment. The geochemical program of5.8,coldacidityof122 mg/L,alkalinityof70 mg/L,and WATEQ4Fversion2.63(BallandNordstrom1991),which net acidity (Eq. 4) of (cid:1)35.9 mg/L as CaCO3 (Fig. 3d, e). isusefultocomputeequilibriumspeciationforlargesample Onceaerationbegan,thepHincreasedandthecoldacidity sets, was used to compute the PCO2, PO2, and saturation decreasedowingtotherapidexsolutionofCO2(Eqs.2,8). index (SI) values for selected minerals. The PCO was Within 12 h, the pH increased to values of 7.0 or more 2 computed on the basis of measured pH, alkalinity, and (Fig. 3d), ultimately attaining a steady-state value of 7.5. temperature. The activities of ferrous and ferric species Protonaciditygeneratedwiththehydrolysisofiron(Eq.1) werecomputedonthebasisofthemeasureddissolvediron, was neutralized by the initial excess alkalinity (Eq. 8), Eh, andtemperature offreshsamples.For consistency, the causing the alkalinity to decline to a steady-state 123 136 MineWaterEnviron(2007)26:128–149 Fig.3 Time-series plots of the quality of water during laboratory acidity,andnetacidity;fmeasuredandestimatedpartialpressureof aerationexperimentontheOttoDischargeconductedDecember3–7, carbon dioxide (PCO ) computed with Eq. 12; g, manganese; 2 2001. a Iron and turbidity; b sulfate and specific conductance; h aluminum; i nickel; j zinc. Values below detection limits were cdissolvedoxygenandtemperature;dpHandEh;ealkalinity,cold plottedasnegative(belowaxis) concentrationofapproximately30 mg/LasCaCO andthe (Fig. 3f). The inverse correlation between pH and PCO 3 2 net acidity to increase to approximately (cid:1)25 mg/L as indicates the potential for PCO to determine pH and, 2 CaCO (Fig. 3e). hence, Fe2+-oxidation rates. To compute PCO as a func- 3 2 The rapid, asymptotic increase in pH coincided with an tion of elapsed time for the aeration experiment, the asymptotic decrease in the PCO from an initial PCO empirical log(PCO ) data were fit with an exponential 2 2 2 value of 10(cid:1)1.1 atm to a steady-state value of 10(cid:1)3.2 atm equation of the form: 123

Description:
Skousen et al. 1998) drainage (AMD) in the upper Schuylkill River Basin, Schuylkill other fish species tend to be limited because of iron-oxide.
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.