7558 Brooklyn Avenue NE

GREGORY L. GLASS Seattle, Washington  98115

Environmental Consultant tel:  (206) 523-1858

______________________________________________________________________________

MEMORANDUM

TO: David South, Ecology/NWRO

SUBJECT: Technical Review Comments, Phase II As-Built Report (May 13, 2009)

UNOCAL Edmonds Site

DATE: July 21, 2009

I have reviewed the Phase II As-Built Report (May 13, 2009).  It effectively documents the second-half cleanup activities at the UNOCAL Edmonds site through Fall 2008.  On behalf of ECAC, I am submitting the following technical review comments.  Please contact me if you have any questions or if further discussions would be useful.

Page 15, Section 4.2.1.1:  the number of steel drums or remnants found in excavation area B1 (18) suggests fairly extensive use of the southeast Lower Yard, where no operational facilities were located, for disposal activities.  Are such steel drums/remnants detectable by surface geophysics (magnetometer)?  To what depth?  If ground water results for southeast Lower Yard wells remain problematic, a survey to detect buried drums that extends past the current treeline may be advisable as a screening investigation.  The extent of contamination uncovered in the southeast Lower Yard may warrant a review of the sediment toxicity test failure in Willow Creek adjacent to this area.

Page 17, Section 4.2.2:  I note that the depth of excavation at area B7, to 5 feet bgs, was not deep enough to address potential slops pond residuals.  While the boring log for MW-510, which was deeper, has no indications of significant contamination being encountered, the subsequent ground water monitoring data strongly indicate residual contamination in this area.

Page 17, Section 4.2.3, second paragraph:  the text states that LNAPL was observed entering the excavation from 2001 backfill material on the east sidewall.  See Figure 4, showing Phase I excavation area B8 to the east.  The text at the top of page 18 identifies sidewall locations removed, which appear to be within the Phase I excavation.  Should the reference to 2001 backfill actually be to 2007 Phase I backfill?  I find it notable that LNAPLs may have entered the Phase II area B8 excavation from clean backfill emplaced just months earlier.  This deserves some additional discussion.

Page 27, Section 7:  the ground water cleanup levels established for the site were based on TPH component data from available ground water monitoring results.  They are best described as interim or default ground water cleanup levels, contingent on TPH composition.  Final cleanup levels will be based on actual TPH composition as measured.  This is important given the changes in TPH composition revealed in the early rounds of the long-term, post-remediation ground water monitoring program.

Page 27, Section 7, second paragraph:  please refer to my comments on the ASI report (comments of April 23, 2009:  Page 12, identification of POC and MNA wells).  The most downgradient wells in the western ground water flow path system are also to be evaluated as POC wells, and MW-509 is also to be evaluated for natural attenuation, as agreed to in our discussions to finalize the Interim Action Work Plan in 2007.  The description of the 40 monitoring wells being sampled therefore should be as follows:  21 POC wells around the perimeter of the Lower Yard, 22 MNA wells including three “plume systems” of 7 wells each plus MW-509, and with 3 wells in both the POC and MNA systems (i.e., 40 = 21 + 22 – 3 “duplicates”).  The symbols on Figure 9 should be revised to show this enumeration of wells.

Page 28, Section 7.1.2:  the summary of well installations in the first paragraph does not account for MW-509 (see Figure 9), which is not in any of the three flow path systems, is not a new POC well, and is not a replacement POC well.  The summation of 20 + 5 + 4 wells equaling the 29 installed wells does not account for the double counting of MW-8R which is a POC replacement well but also is in the western flow path system of wells.  This text should be edited accordingly.  Well MW-509 was installed in the “fourth” LNAPL plume area based on our discussions addressing the final work plans for the Interim Actions, and is to be monitored and evaluated as the other 21 flow path wells are.

Page 28, Section 7.1.2:  the additional monitoring wells continued use of 10-foot screened sections.  Considering the Conceptual Site Model – including horizontal and vertical components of ground water flows, the source(s) and release locations for petroleum contaminants, discharge points to surface water, seasonal water level fluctuations, and so on – could 10-foot screened lengths contribute to some dilution effects for ground water quality as sampled in this manner, biasing the measurements low with respect to possible true impacts on surface water?  If this has been examined and written up previously for this site, please provide a reference.

Page 28, Section 7.1.2:  I note that there are no monitoring wells on the upgradient side (site interior) of the POC wells along the southeast property boundary.  When the long-term, post-Interim Action monitoring well systems were being identified prior to the Interim Actions being performed, the extent of contamination in the southeast Lower Yard was assumed to be less than was actually encountered.  LNAPLs were in fact encountered at excavation area B1, including LNAPLs entering from the upgradient face of the excavation (see page 16).  Depending on further/ongoing ground water monitoring results, the lack of upgradient ground water characterization may be a data gap that needs to be addressed.

Page 29, Section 7.1.2, last paragraph and Section 7.1.3:  the enumeration of wells with respect to collection of soil samples during installation does not address MW-522.  That well appears to have been installed just outside of excavation areas (see Figure 9) and according to the boring log (Appendix J) had a relatively elevated PID reading of 34.2 ppm at the bottom of the boring, with lesser hits above.  No soil samples were taken from the MW-522 boring.  Also, MW-500 and MW-501 both had the notation of slight hydrocarbon-like odor (HCLO) on their boring logs, which could be interpreted as a sign of possible impacts.  Compiling the PID readings and boring log annotations of HCLO versus later ground water chemistry results for TPH (rounds 1 through 4 so far), both indicators during well installations are seen to be variable and imperfect predictors of ground water TPH concentrations.  Nevertheless, it might be useful to at least collect and archive soil samples when elevated PID readings occur or HCLO annotations are made during well installations.

Page 29, Section 7.1.3:  the only notation of “strong HCLO” on well boring logs occurs at MW-129R (Appendix J), and a soil sample for that well showed total TPH of 3,010 ppm.  This correlates well with the post-Interim Action ground water monitoring data for MW-129R, with up to 2,655 ug/L TPH.  The likelihood that the MW-129R boring hit the maximum residual soil contamination in this area is remote.

Page 30, Section 7.1.4:  I note that several wells were redeveloped within about one week prior to the December 2008 round of ground water sampling.  A review of the ground water monitoring results from rounds 1 and 2 shows that TPH concentrations dropped substantially at only two of the seven redeveloped wells (MW-500 and MW-501).  Considering possible future redevelopment actions, perhaps a minimum time delay after redevelopment before sampling should be established as a criterion for the ground water monitoring program.

Page 30 et seq., Section 8 (statistical evaluation):  the statistical evaluation of soils data available from Phase II sampling does not relate to any meaningful spatial unit (e.g., an exposure unit or decision unit) and does not provide a meaningful estimate of an Exposure Point Concentration.  See Figure 7 for disparate areas included in the statistical evaluations of Section 8 (with sampling at the Impacted Stockpile area not shown).  Without a well-defined objective for statistical analysis, these evaluations have limited relevance.  I note that the additional information from Phase II analyses of six samples from monitoring well borings (see Table 6) is not included in the statistical evaluations; the maximum TPH result of 3,010 ppm TPH from MW-129R is higher than the maximum carried forward in the statistical analysis.  As previously noted in my comments on the Phase I As-Built Report, the statistical evaluation of compliance with remediation levels (e.g., soil TPH REL based only on a direct contact scenario) in any case does not address the key question of whether residual soil contamination will cause ongoing ground water quality problems.

Page 34:  The text here says 2,000 tons of contaminated sediments were removed from Willow Creek and 900 tons of clean backfill material was used to restore the creek.  On page 13 the text says original creek topography was restored.  Given the difference in removal and restoration amounts, it is hard to see how the original topography was achieved.  (Could this be a difference in wet/saturated removal amounts versus relatively dry replacement amounts?).  This apparent discrepancy should be explained.

Page 34:  the amount of petroleum impacted soil removed from the Lower Yard is given as 13,759 tons.  This does not appear to match the data provided in Table 2, where the total excavation volume is given as 14,825 tons or 12,825 tons without Willow Creek sediments.  It isn’t clear in either the text on page 34 or in Table 2 how the surficial soils excavated at the asphalt warehouse area that were stockpiled, tested, and then re-used as vadose zone backfill are accounted for – but the text indicates 2 stockpiles were about 250 CY each (see page 20; 500 CY equals 750 tons), which doesn’t seem to resolve the discrepancy in amounts between page 34 text and Table 2.  Table 2 does not include a volume for the approximately 2 feet of additional soil removed at the Impacted Stockpile area (see page 25); could that resolve the apparent discrepancy in volumes of contaminated soils?  That missing volume should probably be added in table 2 in any case.

Page 34:  the last paragraph gives the number of confirmation soil samples collected as 88.  But on page 30 (Section 8.1.1) the statistical analysis data set cites 82 samples, which is the total number included in Table 3.  Is the difference the 6 samples collected during monitoring well installation, as indicated in the well boring logs (Appendix J) and enumerated in Table 6?  If so, it should be noted that the text at the bottom of page 34 does not address the soil REL exceedance in the boring sample at MW-129R.

Table 3:  only one sample with high total TPH results appears to be dominated by TPH-G.  That sample is EX-B1-D-43-4, with TPH-G at 2,000 ppm and TPH-D and TPH-O not detected.  TPH composition will be important for evaluating ground water compliance (as noted elsewhere in these comments), and confidence in the analytical determination of TPH compositions is required. It seems odd that one sample and only one sample from the B1 excavation area would have a dominant TPH-G composition, given the lack of operations in this area and the early date of likely disposal actions.  Assuming the confirmation of the analytical interpretation for TPH composition, can this be explained within the conceptual site model?

Figure 9:  the ground water flow paths as shown do not take into account the influence of the impermeable liner (described in the Phase I As-Built Report; see Figure 10) installed at the southern limit of soil excavations just north of the WSDOT storm drain line.  This feature may be an important consideration for interpreting the results of the ground water monitoring program.  Residual soil contamination near the WSDOT line is at or near the highest levels remaining in the Lower Yard (to the extent currently known).  The liner should be shown on all ground water flow direction figures until it is removed.

Appendix J:  I find it interesting that maximum PID results, and relatively elevated PID measurements, occur at the bottom of the monitoring well borings at a number of locations where wells were installed through clean backfill materials (e.g., MW-512, MW-515, MW-516, MW-520 – and MW-522, apparently just outside of the excavation area).  How do these locations correlate to the locations of elevated residual soil contamination (even if <2,975 ppm TPH, the soil REL) based on confirmational sampling after soil removal?  How many of the locations with elevated residual soil contamination do not have monitoring wells in their vicinity?

Minor edits:

Page 19, Section 4.2.5:  it is worth noting that after the collection of additional samples in this area after Phase I excavations were completed, and before Phase II excavations occurred, the water treatment plant was moved to an area already remediated, providing access to asphalt warehouse area soils.

Page 20 (Section 4.2.5.1) and Table 4:  the text says six samples were collected and analyzed from each of the two stockpiles of asphalt warehouse area soils.  Table 4 provides results for only five samples from each stockpile.  This inconsistency should be resolved.

Page 24, Section 5.2, line 3:  “effected” should be “affected”.

Page 24, Section 5.2, second paragraph:  was the high water mark determined visually (matched to the December 2007 flood event?) or from a 100-year, or similar, flood plain map?

For completeness, it would be good to add a Figure (like Figures 3 through 5) showing the ISP (stockpile area) confirmation samples.

Figure 7:  The box for final excavation quantities at the Railroad Trestle Excavation Area shows 507 CY.  This should be 567 CY (at 1.5 CY per ton; see Table 2 for comparison).

Figure 9:  MW-122 is annotated with “monitoring well was abandoned during the interim action and replaced in October 2008”.  The text does not indicate that MW-122 was abandoned (see Section 7) and the well designation is not given as MW-122R.  Make appropriate changes for consistency.

Appendix A, Figure 2:  The location of the asphalt warehouse as shown is incorrect.