Laserfiche WebLink
1-1, "--'. <br />Fifteen sets r the timc period from 1 <br />of data were r rdautomatically . <br />26, 1993 to July 13, 1993 from the two sites. The results of the testing are presented <br />in Table 1. An Inspection of the raw data reveals an apparent problem with the data <br />from Test testing was conducted too soon after a delivery that too <br />place sometime during the previous evening, Test #7 was a000rdingly was removed <br />from the data base fbr the calculations. : similar problemoccurred for To <br />#14 as well, although the affects are not as large. Test #14 was kept in the data <br />base. <br />The standard deviation for all of the data sets except # 7 is O.o2BO6 gal/hr. <br />This gives a probability i of deteolon forof 0.2 gal/hr <br />when the threshold is set at 'N gaVhr. The corresponding <br />false alarm rate is less <br />than 0.5%. There is no significant bias in the test data. <br />The probability of -r /r a R gal/hr probability o <br />f a <br />false alarm of 5%. The X16ETM will therefore meet the requirements for <br />testing annual <br />tightness as well, <br />The results of this evaluation indicate that the API/Ronan system perform$ <br />adequately on manifolded tanks rto at least jo,000 gallons. The measured <br />performance is as ror■ as is typicallyo+ <br />r ATG systems operating on a single <br />R## gallon • +r - system is notprimarily <br />requirementstester, it does meet the minimum for N, <br />As * used <br />as ed should /.e no problems <br />testing manifolded tanks, particularly formonitoringpurposes. <br />0 <br />L00190018 110 Tialls 6 T92L90TS Xvil OUST %6/9T/6 <br />