SITE INFORMATION AND CORRESPONDENCE

Home Browse Search

m _ I 5 Discussions with laboratory analysts and remediation engine,.-As indicate that it is widely believed chat the greatest number, of false positives will occur when low concentrations (<I00 ug,/L.) This: is su b recent studies b /mobile Oil and S loll 2,0. T MTAE art reed b EPA 80 Y � y Development Company which predict that false positive de=—dons of,Y=E will tend to =ur when <20-50 u&L conccatrarions of NME M reported by EPA method 8020. The data an2lyzed in this report strongly contradict these expectations. The reported EPA=0 values for the 2 false positivesi this data set are ? and 5300 . Therefore,n ..40 r reser tl the reported EPA 8x20 concentration (above 20 ug/L) have no predictive value for indicating the probability of false positire this-identifica(ions. New samples were obLdmed from tfie well previously reporting a false positive EPA 8020 value of 5300 ug/L. Both GC/PM and GC/-MS analysis performed at Lawrence Liverrnore i`ta.tiona l LaboratcrY aJ-NL) confirmed ND levels or,'YME. Our analysis also confirmed the PMsence of damteeethylpentane, a gasoline component which can co--elute with WME.Iead2�o.g to `clue non or ovcreGdn atioth Cd MTBE conc=mtions b EPA A s020. 3-taeshylpeatanc was also independently demoted in these samples by Shell Development Company research labprazories. T7hemforc, this example shows that gasoline comoonenrs present in LUFT ground water,have been mis-idcntined as MTBE by EPA 8020 analysis. Other laboratories have reported that the Pr=ncz of 2rmethylpentane in LUFT ground water samples can result in false dere:ion or MTBE. Therefore, we suggest that analytical laboratories perform an "interference shed:'• analysis of a 3-tnethyipentane and 2-mettiylpentane standards in the presentee and absence of NITBE to evaluate whether this compound co-dotes with NfTBF under routine conditions used in their laboratory for EPA 8020 analysis. It is our opinion that as analytical laboratories opalmize conditions to detect and gaanrfy . fMF by EPA 8020 in thepresence of 3- and?-tnc:hylpentane, both she number and torte.-atiacoa of , reported false positives will decrmse. In addidoncurrent State of Cafifotaia regularioas mare tu use of a single column for analysis of ground wafer samples containing gasoline. The, Association of Caiiiarnia -rest zag Labmawrlcs (ac�jAx3S) has recommended the use of a send canfirtnation column for duction of MTBE in the presence of gasoline to possibly rrdac:, the num be: of false positives detec=d by EPA 8020 (acrtFEV+rS vol. 7, issue 3, fuly 1996). Be..�use Che quality of 12jA 8020 data may improve due to opn timizarions of GC pM separation. conditions and/or the use of a second calumn, we recommend that a comparison of E.n. &020 continue. and 8240 data should be performed in the furore to evaluate whether trends from this data sec Agreement of analytical v2jues reported for EPA 8020 and EPA824Q: For our araIyss, RPD values gmater that: 4051? west unacc=mble, d:monsw- -g a significant di-�xnce berween the values abraiaei by these two analytics/rhe-.}�odz We reviewed elate classes of MISE data sarnalcs with ND or low cowentmdons of gasoline conmounds, sarnples with high concentrations of gasaline and low coueenaltiOns of MTBE, and samples with both Izxgn concentrations of gasoline and MTBE. In gene:-al, we found agreement between the WME cothcenaadons reporr�i by both rne:hcds was fair for the 48 samples that did not contain elevated concenaudons of gasoline carnoounds; 27°,b had Rl?rJ values great=: than 40% (13 of 48 samples containing ND to 7"J ug2.BTrX and/or 7 ug/L B compounds). Interestingly, in the zsha)ority of these cases (9 out & 13), tits ?vi E cance:ncadon reponCd by EPA 8024 was sighhincantly lower L�an the reside 13),daetm-d by EPA $�. We have no explanation as to why EPA 8020 would tend to produce a Dower analytical valve Chan EPA 87so in auprnzim2tely 19% of the samples that do not contain elevated concentrations of gasoline compounds. Only S samples contained cievated conce:htradons of gasoline compounds (BTEX ancJor,3 compounds at concrnaadcas greater or equal to 200 ug/L) and low coat==s or of ylT3E 2