Laserfiche WebLink
December 30, 1994 • <br /> Page 6 <br /> hydrocarbon mixture volatility and mass; and the limitations imposed by mass transfer <br /> mechanisms. Typically, concentration decay continues until an asymptote concentration is <br /> reached. It is reasonable to presume the effectiveness of SVE and air sparging <br /> implementation is exhausted when; (1)the asymptote concentration is reached, (2) the . <br /> asymptote concentration is relatively low compared to initial concentrations, and (3) it is <br /> shown that hydrocarbon impact areas were effected. PACIFIC proposes to terminate <br /> implementation of SVE and air sparging when the effectiveness is lost according to the criteria <br /> mentioned. <br /> SVE influence can be readily documented by collecting subsurface pressure data. <br /> Additionally, an asymptote concentration that does not significantly differ from the initial <br /> concentration infers that areas outside the influence of SVE still exist. Once it has been <br /> shown that influence is as complete as practical, it will be shown that a true asymptote has <br /> been reached. This will be accomplished in two stages. First, pulsed operation of the SVE <br /> will be instituted to show that negligible hydrocarbon mass is available and that the <br /> hydrocarbon mixture lbft in place is relatively immobile. Next, a statistical test will be <br /> completed to show the slope of the hydrocarbon concentration versus time curve is nearly <br /> zero. <br /> The statistical method proposed to document and quantify the asymptotic condition is <br /> published in an American Petroleum Institute(API) publication entitled Technological Limits <br /> of Groundwater Remediation:A Statistical Evaluation Method(Publication Number 4510, <br /> 1991). Generally, the API method defines an asymptotic condition as one in which some <br /> concluding time period in a time series data set exhibits a negligible trend or slope that is not <br /> statistically different than zero. Additionally, both the upper and lower bounds on the slope <br /> (at the 95% confidence interval) should be small. Sequential regression is used to identify <br /> data subsets which are asymptotic. The data subset (concluding time period) which has a <br /> regression slope not different from zero (upper and lower bounds straddle zero) and for which <br /> the absolute value for the lower or upper bound slopes(the largest of the two) is smallest for <br /> all subsets analyzed, is defined as the most trendless data subset. It is proposed that SVE and <br /> air sparging be terminated when a trendless data subset is identified for a time period of <br /> 3 months. <br /> Implementation Schedule <br /> Upon approval of the recommended strategy by Unocal, PACIFIC will submit a scope of <br /> work and fee estimate to initiate remediation. A proposed remediation schedule is presented <br /> below. <br /> 3100011 b/wkpin <br />