GREGORY L. GLASS
7558 Brooklyn Avenue NE
Seattle, Washington 98115
Environmental Consultant tel: (206) 523-1858
DATE: April 19, 2010
TO: David South, Ecology/NWRO
I have reviewed ARCADIS’ February 12, 2010 “2009 Annual Groundwater and Geochemical Parameter Monitoring Report for the Former UNOCAL Edmonds Terminal”. 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.
This report does more to summarize and describe the empirical ground water monitoring results than to evaluate them and further our understanding of site ground water contamination. The extended text summaries of the monitoring results for each of the first 7 rounds of sampling would be better presented as tables. As an illustration, I have drafted 3 summary tables of the ground water TPH results for rounds 1 through 7 (see separate .xls file, with 3 worksheets). The first table lists the number of wells exceeding the default ground water TPH CULs (east and west Lower Yard) and the ranges of those exceedances, in ug/L, for POC and Interior wells by sampling round. The second and third tables summarize the number of TPH default CUL exceedances, by well (POC wells on the second table and Interior wells on the third table), with the range of those exceedances noted. The ordering of wells in these two tables follows the spatial layout of the wells, making it easier to visualize the results across the entire Lower Yard. Tables of this type, which summarize all of the TPH results for the first 7 sampling rounds “on a page” (by round, or by well), highlight several features of the ground water monitoring results to date; they tell the story better than extended text summaries.
The final paragraph in the Conclusions section (Section 7, page 31) states: “Overall, the groundwater IHS concentrations at the site have remained below the specified CULs at most monitoring locations following the interim actions.” This statement is not supported by the data. As shown in the attached tables, 25 out of 40 wells have had exceedances of the default TPH CULs over the first 7 sampling rounds: 14 of 21 POC wells (67%), and 11 of 19 Interior wells (58%). Almost half of the POC wells (10 of 21, or 48%) have had TPH exceedances for 4 or more (i.e., the majority) of the first 7 sampling rounds. In the last of the 7 rounds included in this report, in October 2009, 11 of 21 POC wells (52%) still exceeded for TPH. The data summaries presented in the attached tables also show that TPH contamination overall has been higher, and more persistent, in the POC wells compared to the Interior wells.
TPH cleanup levels (CULs) are determined based on the composition of TPH types in a ground water sample. The CULs that are cited in this report (e.g., see Table 1), and that have been used in reporting on each sampling round as results were submitted to Ecology, are based on average historic compositions for portions of the Lower Yard; they are in reality default CULs (a point I have noted in previous comments). Final CULs and compliance evaluations should be based on the actual TPH composition of ground water samples.
In that regard, it is interesting to note the effect that a change in analytical labs (see Section 3.3, page 7, bottom paragraph) before the February 2009 sampling round had on the reported TPH composition for multiple wells, and particularly the shift toward higher TPH-D contributions. CULs will shift lower as the TPH-D component becomes more dominant. The change in analytical labs achieved a primary goal of lower TPH detection limits. The apparent change in TPH composition associated with this change in labs is not noted in the report but is worthy of discussion with respect to causes and implications for CULs.
Recent re-evaluations of ground water/surface water interactions at the site, and measured ground water elevation data sets, have led to some revisions in the model for ground water flows at the site. Ground water contour maps are provided; a discussion of the revised model for ground water/surface water interactions would also be very useful, to update the site conceptual model for ground water. A number of measured ground water elevations have been determined to be questionable (anomalous) and are not used in developing ground water contours; they are so identified on the contour plots. A discussion of the possible causes for such anomalous water elevation measurements, and corrective actions to minimize their future occurrence, would be a useful addition to the report. A similar discussion of anomalous measurements of field parameters would also be useful.
The ground water monitoring system, with 40 wells being sampled for chemical analyses, consists of a mixture of older wells and wells installed just prior to the first sampling round in October 2008. Were all 40 wells developed or redeveloped before the first round of sampling? It is interesting to note that the initial results for TPH were anomalously low at a number of wells (e.g., MW-129R, a new well, and MW-136, an older well) considering the total record over the first 7 rounds of sampling. Inclusion of such low initial TPH results in time trend analyses can have a substantial impact on the results (as would the inclusion of any “anomalous” data). It is of interest to consider whether the initial monitoring data are representative of “equilibrium” conditions, with respect to soil excavations, ground water removal during excavations, and well development activities.
The time trend analyses use log-scaled concentrations and linear time. Such log-linear regressions are equivalent to an exponential decay model for concentrations over time. How was time of sampling entered into the regressions? What time is represented by the y-intercept value (i.e., what is time zero)? The fitted curves would support an estimate of the time-to-meet-CULs for any decreasing concentration trends; that information is not explicitly presented in the report. (Note that such time estimates can be highly dependent on the chosen model, and can differ markedly for regressions using linear concentration versus log-concentration values). The time-to-meet-CULs is not meaningful for cases of increasing time trends for concentration. With respect to the status of site ground water and the likelihood of meeting CULs at the end of the two-year monitoring program, however, the mere occurrence of increasing concentration trends is notable at this point in the monitoring program.
The period of sampling available for this initial time trends evaluation is obviously limited. It should be noted that it covers sampling rounds from October to October, and thus includes somewhat more than one annual hydrologic cycle. This could be important if there is seasonality in ground water contaminant response (requiring a seasonal trends analysis). With the limited available data set, it could be interesting to see if there is a general systemic response (i.e., across most wells with more than de minimis TPH concentrations) to changes in ground water levels, which could imply a seasonality component to trends. Plotting TPH data for multiple wells for the ECAC blog, I have noted that multiple wells seem to show similar up-and-down patterns in TPH over sampling rounds.
The development of a ground water mound in the southeast Lower Yard after completion of soil excavations and backfilling is noted. The potential significance (including vertical gradients) of that ground water mound for TPH concentrations measured in the two most directly downgradient wells, MW-500 and MW-501, is one aspect of the relationship between ground water elevations and ground water chemistry that should be discussed.
In contrast to relatively simple models of steadily decreasing concentrations of TPH over time, as a result of natural attenuation processes, some wells at this site have shown a pattern of highly variable and “jumpy” TPH concentrations over time. Such patterns include exceedances of TPH default CULs after one or more rounds of concentrations well below CULs. These variable patterns are of even more interest assuming source materials have been removed through the extensive Lower Yard Interim Actions. Given variable TPH patterns over time at some wells (see, for example, MW-104, MW-143, LM-2, MW-135, MW-136 MW-500, MW-501, MW-502, MW-507), time trend analyses will be sensitive to the exact time periods included in the analysis and may have quite limited predictive power. To give but one example, the time trend analysis for MW-104 (Appendix I) gives a (non-statistically significant) decreasing trend, but the round 8 TPH concentration of 3,105 ug/L substantially exceeds any result from rounds 1 through 7. The presentation of time trends in this report would be improved by additional discussion of these characteristics of the ground water data set over the first 7 rounds, and cautions against over-reliance on the predictive validity of the results.
There are a number of monitoring wells where reported TPH concentrations are all, or nearly all, not-detected for the first 7 sampling rounds (e.g., MW-109, MW-511, and MW-524). In the discussion of natural attenuation (geochemical) parameters in Section 5, it would be interesting to compare the geochemical profiles for these wells with de minimis levels of TPH to the profiles at other wells with higher TPH concentrations in support of the interpretation of ground water geochemistry.
Minor Edits
The date on the draft report cover shows 2009, but should be 2010.
Section 3.1, page 4. The list of wells included as interior monitoring wells (n=19), at the bottom of the page, actually lists 20 wells. MW-518, a POC well included in the list of POC wells on page 4, is included under interior monitoring wells in error and should be deleted.
Section 3.3, page 6. The 56 on-site wells noted in the first sentence includes 8 piezometers that were installed in late July 2009 (see Section 3.4.6.1) and therefore were only available for gauging starting in August 2009. Since the text in Section 3.4 identifies 49 wells being gauged prior to the installation of those piezometers, it would avoid potential confusion if the text in the first paragraph of Section 3.3 distinguished between the 49 wells (including one offsite well, MW-301) and the 8 piezometers, and noted when the gauging program expanded from 49 to 57 locations.
Section 3.3, page 7. At the end of the top paragraph, a statement can be added to note that the time for collecting water level information at each well in each round is listed in Table 2, and the total time for completing a round of water elevation measurements can therefore be determined from the data in Table 2.
Section 3.3, page 7. The bottom paragraph says samples were sent to a new analytical lab (Lancaster Laboratories) starting with the January 2009 event. The next event after December 2008 was not in January 2009 but in February 2009 (see Section 3.4.3). This should be corrected in the text. (Perhaps January 2009 reflects the time for completion of making new lab arrangements, not the date for the first sampling event when the new lab was used?)
Section 3.4, page 8. In the first line of the second paragraph, “grounding” should be changed to “mounding”.
Section 3.4.8, page 13. The LNAPL thickness reported as “0.01 inch” should be “0.01 feet”. See Table 2 – LNAPL (apparent) thicknesses are given in units of feet. As the sentence on page 13 notes, the LNAPL thickness at MW-510 in October 2009 was measured from 6.71 feet to 6.72 feet amsl.
Section 3.6.2, page 17 (subsection 3.6.2.2). For the December 2008 sampling round summary, the text notes 5 interior wells above TPH CULs, 6 with detectable TPH below CULs, and 9 below the TPH detection limits. Those 3 categories sum to 20 interior wells instead of 19. Revise the counts as necessary.
Section 3.6.3.1, page 18. The counts for POC wells appear to be in error and should be checked. There are 10 wells with TPH above CULs and 8 with TPH above detection limits but below CULs.
Section 3.6.4.1, page 20. The number of POC wells is mis-stated as 19; there are 21 POC wells. The well counts in this subsection should be checked to be sure all 21 POC wells are included.
Section 3.6.6.1, page 23. In the second paragraph the number of POC wells is mis-stated as 19; there are 21 POC wells.
Section 6, page29. On the third line from the bottom, delete the second redundant “were”.
The symbols used for monitoring wells on figures appear to have several errors that should be corrected. The list of POC wells (n=21) is given in Section 3.1, Monitoring Well Network (page 4). The symbols for POC wells and other, non-POC wells are not correct in some cases; for example, on Figure 18, MW-151 (a non-POC well, see Section 3.1) is shown with a POC symbol, MW-129R (a POC well, see Section 3.1) is shown with a non-POC symbol, and MW-523 (a POC well, see Section 3.1) is shown with a non-POC symbol. All wells on all figures were not checked for our review, but corrections across all figures should be made as needed.
UNOCAL Edmonds Former Bulk Fuel Terminal
Summary of Ground Water Monitoring Results
Round 1 through Round 7 (Oct 2008 - Oct 2009)
TPH Exceedances of Default Ground Water Cleanup Levels
POC Wells [n=21] Interior Wells [n=19]
number low high max location number low high
Sampling Round (ug/L) (ug/L) (ug/L) (ug/L)
Oct 2008 6 909 8330 MW-501 9 701 1697
Dec 2008 4 968 5407 MW-510 5 562 1438
Feb 2009 10 595 16380 MW-510 6 748 2005
Apr 2009 9 779 22930 MW-510 6 534 [533] 1503
June 2009 9 915 25090 MW-510 5 512 [413] 1175
Aug 2009 7 1066 18080 MW-510 3 539 [561] 847 [755]
Oct 2009 11 715 2,720 (a) MW-147 (a) 4 534 [594] 1013
NOTES
(a) LNAPLs detected at MW-510, no sample collected for lab analysis
[533] A second TPH concentration in brackets is for a field duplicate sample
UNOCAL Edmonds Former Bulk Fuel Terminal
Summary of Ground Water Monitoring Results: POC Wells
Round 1 through Round 7 (Oct 2008 - Oct 2009)
TPH Exceedances of Default Ground Water Cleanup Levels
number low high
(ug/L) (ug/L)
POC Well [n=21] (a)
MW-150 1 715
MW-149R 1 775
MW-524 0
MW-147 5 968 2720
MW-523 0
MW-8R 0
MW-522 3 779 955
MW-20R 2 866 1087
MW-104 4 916 1,915 [2,046]
MW-101 0
MW-518 6 1007 1403
MW-139R 0
MW-510 (b) 7 3980 25,090 (b)
LM-2 4 915 2225
MW-108 0
MW-109 0
MW-129R 5 1675 4425
MW-501 4 815 [785] 8300
MW-500 5 595 3000
MW-135 4 1695 2625
MW-136 5 1622 3725
NOTES
(a) POC wells are listed in spatial order, from southwest to southeast site boundary
(b) LNAPLs detected in October 2009 at MW-510, no sample collected for lab analysis
[2,046] A second TPH concentration in brackets is for a field duplicate sample
UNOCAL Edmonds Former Bulk Fuel Terminal
Summary of Ground Water Monitoring Results: Interior Wells
Round 1 through Round 7 (Oct 2008 - Oct 2009)
TPH Exceedances of Default Ground Water Cleanup Levels
number low high
(ug/L) (ug/L)
Interior Well [n=19] (a)
MW-143 3 770 2005
MW-519 0
MW-520 1 731
MW-521 0
MW-511 0
MW-512 4 534 [533] 748
MW-513 6 534 [594] 932
MW-514 7 825 1578
MW-515 1 947 [975]
MW-516 1 801
MW-517 0
MW-502 6 512 [413] 1700
MW-503 0
MW-504 1 701
MW-505 0
MW-506 0
MW-507 7 562 1142
MW-508 1 745 [645]
MW-509 0
NOTES
(a) Interior wells are listed in approximate spatial order, from southwest to southeast of Lower Yard
(in original MNA plume layout, west-central-east plumes, from upgradient to downgradient)
[533] A second TPH concentration in brackets is for a field duplicate sample
Comments