[Home]

Meeting Minutes
Facility Investigations, Closure Projects,
and Well Installations
Camp Stanley Storage Activity
F11623-94-D0024/RL83
Parsons ES 736071.02

Date:             22 May 2001

Time:            9:00 A.M. - 4:30 P.M.

Place:            Camp Stanley Storage Activity, Boerne, Texas

Subject:         CSSA Data Quality Objectives for Groundwater Monitoring Program

         Attendees:    

Minutes prepared by Julie Burdey, Parsons ES.

The meeting commenced at 9:00 AM at the CSSA conference room.  The objectives of the meeting were to 1) continue the discussion of DQOs for the upcoming June 2001 quarterly groundwater monitoring event; and 2) discuss draft schedule for all delivery orders currently being prepared by Parsons ES. 

DQO Discussion

The meeting began with a review of the preliminary draft RL74 TIM #5 minutes.  The first activity was to identify items that were discussed during that meeting that pertained to the June 2001 quarterly groundwater monitoring event.  Of the general CSM points, it was agreed that items 3, 4, 5, and 7 pertained to the upcoming monitoring.  Objectives of the groundwater program that pertain to the monitoring include: monitoring groundwater, evaluating open versus closed boreholes, evaluating well placement, identifying sources of contamination, determining lateral and vertical extent of contamination, determining flow direction and seasonal variability, comparing formations, updating/refining conceptual site model, evaluating need for additional transducers, and obtaining pumping data from other local sources.  Associated unknowns include lateral/vertical extent of contamination, open/cased boreholes (effect on contamination levels), dilution, local flow and direction, contaminants of concern in southwest corner, number of plumes, trends, and metals contamination/background.

Next, the group continued going through the DQO process for CSSA groundwater monitoring, and the results are presented in Attachment 3.

Based on the DQO discussion, several items which should potentially be considered for the upcoming well installation project were identified.  These include:

·        Is there any need for additional Bexar Shale wells?

·        Discrete interval grab sampling during drilling of the RL83 wells has provided much information that would otherwise have been missed based on the screen placement at the bottom of the formation.

·        Should installation of additional shallow Lower Glen Rose (Glen Rose cluster wells) wells be added to the new project?

The most resource-effective monitoring design that satisfies all of the DQOs was identified at the end of the discussion.  For the June 2001 groundwater monitoring event, each of the wells identified in Attachment 2 will be sampled.  Samples from the existing monitoring wells will be analyzed for metals and the short list of VOCs approved by EPA.  Samples from drinking water wells and the new cluster wells will be analyzed for the full list of VOCs (plus acetone) and metals.  All data will be validated in accordance with AFCEE QAPP requirements.  Any rejected data will be brought to CSSA and AFCEE’s attention as soon as possible so that the need for resampling can be evaluated.  The quality of drinking water well data will be required to be very high and results flagged “R” will require resampling; more flexibility is allowed for monitoring well data.  Drinking water well data flagged with “F”, indicating that the value is between the MDL and RL, should not require resampling as all the RLs are below the MCLs.  Drinking water well data flagged with the "J" qualifier, which is added for a variety of reasons associated with non-compliant quality control data (e.g. non-compliant  surrogate recoveries (detected values), non-compliant

laboratory control sample recoveries (detected values)), will be considered for resampling on a case by case basis.    If any drinking water well results are within the 90-110% MCL or TNRCC action level range, that well will be resampled for the parameters in that range.  Due to the large amount of historic data from the existing wells, duplicate samples will not be collected from the existing well group.  Duplicates will be collected from the new wells only, at a rate of one per ten samples.  MS samples will be collected at a rate of one per twenty samples (including all wells).  MSD samples will not be collected.  The need for 100% data validation will be evaluated for future monitoring events.

Schedule

Two schedules were discussed:  1) AFCEE and CSSA’s internal schedule for analytical package review and 2) schedule prepared by Parsons ES for deliverables on all projects.

For the analytical package groups, CSSA stated that the deliverable status table that Parsons previously used was the most useful for CSSA and AFCEE.  Parsons ES agreed to update it with the most recent information and forward it to CSSA, AFCEE, and Informatics for use in setting up their internal tracking system.

Several schedules prepared by Ed Strayer were distributed. 

·           CSSA asked that 3-month projections be given weekly.  The window should start with the present date, and just show the future. 

·           Parsons actions should be shown in blue, and CSSA/AFCEE actions shown in green. 

·           Details should be provided every month, and a rolled up schedule submitted every two weeks. 

·           PIMS actions should be added to the schedule. 

·           A task for each item titled “Incorporate into Electronic Encyclopedia” will be added. 

·           For appropriate items, a task titled “Add data into GIS-SDS” will also be added. 

·           For deliverables, an item titled “CSSA/AFCEE Approve/Reject Response to Comments” will also be added. 

·           CSSA would like each update to populate their CSSA Outlook team calendar automatically, and delete old entries automatically. 

Parsons ES suggested that draft technical reports, especially those for RL17, begin to be submitted so that the contract performance period schedule requirements can be met.  CSSA agreed that Parsons could start submitting these reports, even with associated analytical packages still pending approval. 

Attachment 1  EPA DQO Process

Attachment 2  Wells to be Sampled, June 2001

Attachment 3  DQO Evaluation

Objective:  Determine the Vertical and Lateral Extent of Groundwater Contamination

Step one:  State the Problem

1.     Members of Scoping Team.

CSSA, AFCEE, WPI, Informatics, and Parsons ES

2.     Develop/refine the CSM.

Chlorinated solvent groundwater contamination is known to exist in the Lower Glen Rose formation water-bearing unit in the vicinity of well 16 (central portion of CSSA) and in the southwest corner of CSSA.  The lateral extent of contamination in these areas is unknown.  It is thought that these two areas may represent two separate plumes.  Low levels of chlorinated solvents have also been recently detected at CSSA Well 1 (south of CSSA).  Since most of the wells in which contamination has been detected cross multiple water-bearing zones, the vertical extent of contamination is also unknown.

3.     Define exposure pathways and exposure scenarios.

Exposure could potentially occur at CSSA water supply wells 1, 9, and 10, and at off-site well LS-7.  Additional exposure pathways are probable due to the large number of water supply wells (both private and municipal) in the area; however, specifics regarding contamination levels in these wells are currently unavailable. 

4.     Identify available resources.

Groundwater monitoring is planned to be conducted under AETC delivery order DO5084; however, a modification to the contract will be necessary.  Funds initially intended for monitoring of 20 off-site wells will be reprogrammed for sampling of 24 wells identified in Attachment 2.  A laboratory capable of attaining quality requirements specified in the AFCEE QAPP has been identified and audited for several SW-846 methods.

5.     Write a brief summary of the contamination problem.

See item 2 above.

Step Two:  Identify the Decision

1.     Identify the key decision.

Is the well network adequate to define the extent of contamination?

Other associated decisions/questions include:  How many monitoring events are required to determine the objective?  Are any contaminants occurring in new wells?  What are the contaminant-specific action levels and requirements?

2.     Identify alternative actions that may be taken based on the findings of the field investigation.

None noted.

3.     Identify relationships between this decision and any other current or subsequent decisions.

See Step One, item 1.

Step Three:  Identify the Inputs to the Decision

1.     Identify the information inputs needed to resolve the decision.

Analytical data to define the COCs and action levels to define extent. 

2.     Identify sources for each information input, and list those inputs that are obtained through environmental measurements.

Source of analytical data will be groundwater samples collected during groundwater monitoring program. 

3.     Define the basis for establishing contaminant-specific action levels.

To the extent possible, action levels are defined by regulations.  Data from drinking water wells will be compared to MCLs and Texas Action Levels.  For drinking water wells, the action will be to treat water at the wellhead, provide an alternate source of water, or remove the well from the distribution system.  For all other wells, the reporting limit (RL) is considered the action level, and for these wells, the action is continued monitoring.

4.     Identify potential sampling approaches and appropriate analytical methods.

Sampling approaches depend on the construction of the well.  For the June 2001 event, sampling approaches are listed in Attachment 2.  SW-846 analytical methods are planned for the June 2001 event.  As noted in Attachment 2, samples from new wells and LS-7 will be analyzed for the entire list of SW-8260A analytes.  Existing wells will be analyzed for the short list only.  500 series may be appropriate for drinking water wells, though use of it has never been required by regulatory agencies.  Samples will also be analyzed for the metals arsenic, barium, cadmium, chromium, copper, lead, mercury, nickel, and zinc.

Step Four:  Define the Boundaries of the Study

1.     Define the geographic areas of the field investigation.

The geographic area of the field investigation is the CSSA facility and the offsite area southwest of CSSA.

2.     Define each environmental medium of concern.

Groundwater.

3.     Divide each medium into strata having relatively homogeneous characteristics.

Lower Glen Rose groundwater, Bexar Shale groundwater (if any), and Cow Creek groundwater.

4.     Define the scale of decision making.

Temporally – decisions regarding the well network can be made on a quarterly basis.  Spatially – above-action level analytical data should be bounded by non-detect analytical data to define the extent of contamination, both vertically and laterally.  However, analytical data very close to the action level could also be used to define the extent.

5.     Determine the time frame to which the decision applies.

See Step Four, item 4 above.

6.     Determine when to take samples.

EPA 3008(h) order requires quarterly monitoring.  This interval is sufficient for acquiring seasonal data.  Occasionally, CSSA may consider collecting samples after heavy precipitation since a correlation between increased contaminant levels and precipitation has been observed.

7.     Identify practical constraints that may hinder sample collection.

Sample locations are dictated by locations of monitoring wells.  Installation of monitoring wells is expensive and can take several months, depending on the number of wells.

Step Five:  Develop a Decision Rule

1.     Specify the parameter of interest.

The maximum concentration of each analyte is the parameter of interest.

2.     Specify the action level for the decision.

For monitoring wells, the action level is the reporting limit.  For drinking water wells, the action level is the MCL or the Texas action level, whichever is lower.  If the result is below these action levels, then the monitoring network will be considered adequate in that area.  If the result is above action levels, then the well network may not be adequate in that area and additional wells may be necessary.

3.     Combine the outputs into an “if,… then” decision rule.

If the result is below these action levels, then the monitoring network will be considered adequate in that area.  If the result is above action levels, then the well network may not be adequate in that area and additional wells may be necessary.

Step Six:  Specify Limits on Decision Errors

1.     Determine the possible range of the parameter of interest

Reporting limits must be equal to, but preferably below, action limits identified above.  Highest levels detected are listed below:

2.     Define both types of decision errors and identify the potential consequences of each.

Error one (false negative, false nondetect results):  well network is mistakenly considered to be adequate in a given area.  Potential consequence:  For any one monitoring event, the consequence for CSSA monitoring wells is a delayed identification of the plume location; the well network will be re-evaluated quarterly.  The monitoring wells at CSSA are not being used for any other purpose (such as water supply, irrigation, etc).  However, for drinking water wells, the consequence is potential ingestion of above-MCL water by humans.

Error two (false positive, false above action-level results):  well network is mistakenly considered to be inadequate in a given area.  Potential consequence:  Unnecessary additional wells could be installed in some locations.  Well head treatment or supply of an alternate source of water could be initiated unnecessarily.

3.     Specify a range of possible parameter values where the consequences of decision errors are relatively minor (gray region).

Detected results near the action level would result in a gray area.  Also, a one-time detection of an analyte above the action level, followed be several rounds of nondetect could be considered a gray area.

4.     Assign probability values to points above and below the action level that reflect the acceptable probability for the occurrence of decision errors.

Assignment of quantitative probability values associated with CSSA’s acceptable limits for making an incorrect decision is not appropriate based on the number of factors involved.  Decision errors associated with drinking water wells are very important to avoid.  Decision errors are most likely when detected contaminant levels are at or near the MCL.  During the data validation process, a review will be made to ensure that drinking water well sampling and analysis are in 100% concurrence with the AFCEE QAPP and the approved variances.  Any deviations will be identified to CSSA and AFCEE as soon as possible so that an appropriate corrective action can be identified.  In these cases, resampling may be necessary.  In addition, if detected analytical results for the drinking water wells are within 10% of the MCL, resampling of the wells where the detection occurred will take place.  Resampling will only be for the analytes within the 10% range, unless there were other QA/QC problems requiring resampling.

For all wells (drinking water and monitoring), results will be evaluated against historic data, where available.  Any results which do not agree with previous trends will also be carefully evaluated to ensure that it is compliant with the AFCEE QAPP (and approved variances).  Where discrepancies are identified, CSSA and AFCEE will be notified as soon as possible so that an appropriate corrective action can be identified.

*  According to 30 TAC 290.107(c)(2)(C), if VOC contaminant is detected at a level exceeding 0.5 μg/L in any sample, then the system must monitor quarterly at each point of entry to the distribution system which resulted in a detection.

5.     Check the limits on decision errors to ensure that they accurately reflect the decision-maker’s concern about the relative consequences for each type of decision error.

The acceptable limits on decision errors are smallest for cases where there is greatest concern for decision errors – incorrectly identifying wells as exceeding drinking water standards, and not identifying drinking water wells that do exceed drinking water standards.

Step Seven:  Optimize the Design

1.  Review the DQO outputs and existing environmental data.

See Steps 1-6.

2.  Develop general sampling and analysis design alternatives.

Sampling locations are limited to well locations.  Ultimately, each of the CSSA wells will be sampled via a low-flow pump; however, these pumps will not all be installed by the June 2001 sampling event.  Private wells are sampled at the tap.  Samples will be analyzed via SW-846 methods for consistency in analytical results.  Level of data validation and quality of data provide design alternatives. 

One alternative is that all analytical data be required to meet stringent AFCEE QAPP criteria and undergo 100% data validation.  Data resulting from this level of QA/QC could be used for site characterization, risk assessment, and identification of remedial alternatives.  However, a large amount of very high quality data is already available for many of the wells.  Producing data with this very high quality level costs more in data validation time, but does not necessarily provide new information.

An alternative is to require very high quality data for all drinking water wells and new wells.  Data from the other wells does not need data validation, nor does it need to meet stringent AFCEE QAPP requirements.  This alternative will be evaluated for future monitoring events.

Other alternatives are associated with the number and type of QA samples needed. In the past, duplicate groundwater samples were collected during every sampling event, at a rate of one after every nine environmental samples.  MS and MSD samples were collected during every sampling event, at a rate of one MS and one MSD after every eighteen samples (including duplicates).  These collection frequencies were based on a conservative interpretation of the AFCEE QAPP.

Since duplicates have been collected for approximately five years of monitoring and no discrepancies between the sample and duplicate results have been identified, an alternative is to limit collection of duplicate samples only to new wells that have not previously been sampled.  This would allow comparison of duplicate results collected from new pumps.  In addition, the duplicates could be collected at a rate of one per ten samples, the alternative, commonly accepted interpretation of the AFCEE QAPP.

An alternative for MS/MSD samples is the collection and analysis of a matrix spike (MS) and a matrix duplicate (MD).  The matrix spike is used to assess the accuracy of the method in a given sample matrix.  Previously, a matrix spike duplicate has been used to assess the precision of the method in a given sample matrix.  However, the analysis of an MSD also provides accuracy information because the sample is spiked with a known concentration of the target analyte(s).  The guidance in SW-846 indicates that a matrix duplicate may be used to assess matrix precision instead of using a matrix spike duplicate.  A matrix duplicate is defined as “An intralaboratory split sample which is used to document the precision of a method in a given sample matrix” and is also commonly referred to as a laboratory duplicate or an analytical duplicate.  The use of an MS/MD prevents confusion that may arise when the accuracy of the MS and the accuracy of the MSD are in conflict because an MD can only be used to assess precision.  An alternative for the collection frequency of the MS/MD is to collect one MS and one MD per twenty normal field samples.  This collection frequency is based on the alternative, commonly accepted interpretation of the AFCEE QAPP.

3.  For each design alternative, verify that the DQOs are satisfied.

See Step 7 Item 2.

4.  Select the most resource-effective design that satisfies all of the DQOs.

For the June 2001 groundwater monitoring event, each of the wells identified in Attachment 2 of the RL83 TIM #4 Minutes will be sampled.  Samples from the existing monitoring wells will be analyzed for metals and the short list of VOCs approved by EPA.  Samples from drinking water wells and the new cluster wells will be analyzed for the full list of VOCs (plus acetone) and metals.  All data will be validated in accordance with AFCEE QAPP requirements.  Any rejected data will be brought to CSSA and AFCEE’s attention as soon as possible so that the need for resampling can be evaluated.  The quality of drinking water well data will be required to be very high and results flagged “R” will require resampling; more flexibility is allowed for monitoring well data.  Drinking water well data flagged with “F”, indicating that the value is between the MDL and RL, should not require resampling as all the RLs are below the MCLs.  Drinking water well data flagged with the "J" qualifier, which is added for a variety of reasons associated with non-compliant quality control data (e.g. non-compliant  surrogate recoveries (detected values), non-compliant laboratory control sample recoveries (detected values)), will be considered for resampling on a case by case basis.If any drinking water well results are within the 90-110% MCL or TNRCC action level range, that well will be resampled for the parameters in that range.  Due to the large amount of historic data from the existing wells, duplicate samples will not be collected from the existing well group.  Duplicates will be collected from the new wells only, at a rate of one per ten samples.  MS samples will be collected at a rate of one per twenty samples (including all wells).  MSD samples will not be collected.  The need for 100% data validation will be evaluated for future monitoring events.

Attachment 4  Excerpt from SW-846 regarding MS/MSD samples

Attachment 5  Chemistry Review Process