Tannery sludge containing chromium VI (Cr6) was applied as fertilizer on over 100 farm fields in four counties in northwest Missouri from 1983 to 2009. The Missouri Department of Natural Resources (MDNR), in partnership with the Missouri Department of Health and Senior Services (MDHSS) and the U.S. Environmental Protection Agency (EPA), established site-specific screening levels and decision units (DUs) for the purpose of designing and conducting a pre- Comprehensive Environmental Response, Compensation and Liability Information System (CERCLIS) screening investigation of the affected agricultural fields over two soil sampling events in 2009 and 2010.
Through a collaborative effort, a plan was developed to meet the challenges of scale by utilizing real-time field portable X-ray fluorescence (FP-XRF) technology and real-time statistical assessment of the XRF data using the Visual Sample Plan (VSP) software. The purpose of the real-time data assessment was to ensure that statistically adequate sampling of affected fields was accomplished during a single mobilization. The ability for the tools to accomplish this was demonstrated in a pilot study where data collection and assessment was performed in the field. Originally, a field-based dynamic work strategy (DWS)-based sampling design was also planned for the main study. However, the field aspects of the project had to be modified at the last minute due to frozen ground. Soil samples had to be chiseled out of the ground and brought inside a building to thaw and dry prior to sample processing. XRF analysis on the processed samples and real-time data evaluation was performed inside as well. Despite the change of plans, the investigation achieved adequate characterization of the affected lands while saving significant time and money. Results from the investigation concluded that the historic use of tannery wastewater treatment sludge as a farm field fertilizer does not pose a current health threat to farmers working the fields or to residents living adjacent to sludge-treated fields.
|Site Name||Tannery Sludge Farm Fields: Agricultural Soil Analysis Phase|
|Location||Andrew, Buchanan, Clinton, and DeKalb Counties, MO|
|Site Type||Agriculture Applications|
|Project Lead Organization||MDNR Hazardous Waste Program (HWP)|
|Project Lead Type||State Lead|
|Regulatory Lead Program||Pre-CERCLIS Screening|
|Reuse Objective Identified||Yes|
|Proposed Reuse:||Other (agricultural)|
Wastewater treatment sludge from the Prime Tanning / National Beef Leathers (NBL) leather tannery in St. Joseph, Missouri, was distributed and applied as an agricultural fertilizer on more than 100 farm fields across four counties in northwest Missouri from 1983 to 2009. The tannery sludge contained Cr6, a naturally occurring metal that can be carcinogenic and poses a significant health risk. With several thousand acres affected, the MDNR, working with the MDHSS and the EPA, began a comprehensive investigation into the impacts of tannery sludge on farm fields in Andrew, Buchanan, Clinton, and DeKalb counties in northwest Missouri.
The purpose of the investigation was to conduct a Pre-CERCLIS site screening investigation to determine whether there has been a release of hazardous substance into the environment that could present a danger to human health or the environment. Given the size and number of the farm fields affected, a sampling effort was designed using incremental composite sampling (ICS), real-time analysis of a proxy, and a combination of statistical and worst case sample selection to determine sample numbers.
An adaptive sampling strategy utilizing real-time measurement technologies saved significant time and money by reducing fixed-base laboratory costs and turnaround times. The DMA led to the choice for a more sophisticated analytical method, avoiding the waste of generating unusable data results. The sampling results from 19 worst-case farm fields showed that concentrations of Cr6 exceeded empirically derived background concentrations by a factor of three, but did not exceed the site-specific health-based screening levels established by the EPA Region 7 risk assessment staff, the Agency for Toxic Substances and Disease Registry (ATSDR), and the MDHSS. This conclusion was supported by data gathered from nearly 600 soil samples analyzed using an XRF instrument, and 100 Cr6 analyses by a fixed-base laboratory. The highest Cr6 result observed was more than an order of magnitude below the farm-field screening level. In conclusion, tannery wastewater treatment sludge does not pose a current health threat to residents or farmers living on or near sludge-treated agricultural areas.
The planning team developed the conceptual site model (CSM), established risk-based screening levels, and designed and implemented a DMA/pilot study prior to creating a sampling and analysis plan (SAP) for the pre-CERCLIS site screening investigation. XRF reports total chromium and cannot distinguish between different species of chromium (i.e., Cr3 vs. Cr6). One identified data gap was the lack of knowledge regarding a reasonably stable relationship between total chromium and Cr6 concentrations. If there was a stable relationship between concentrations, the variability in the soil’s total chromium results (from real-time XRF analysis) could be used to conservatively predict the variability in soil Cr6 concentrations. Being able to predict analyte variability is required for confident statistical prediction of sample numbers. Otherwise, the project team would have to wait until the Cr6 data came back from the laboratory to know whether enough data had been gathered. The CSM identified the sludge application technique and fate mechanisms for release of Cr6 from the sludge, potential exposure pathways, other possible sources, the chemical composition of the sludge waste at the tannery, and the need for site-specific Cr6 background concentrations. The DMA: 1) established the sampling design that would produce data that were representative and statistically significant, 2) established the sample processing/homogenization procedures, 3) demonstrated the applicability of XRF technology for using variability in total chromium concentration as a proxy for Cr6 variability, 4) showed that total chromium concentrations could be used to predict ball-park estimates of Cr6 concentrations, 5) provided preliminary evidence that soil Cr6 concentrations were low, and 6) determined which analytical method could produce usable Cr6 data from this matrix. The site-specific screening level for Cr6 was 86 milligrams per kilograms (mg/kg) and was developed specifically for farm fields. The exposure pathway that drove the farm field screening level was inhalation of particulates by farm families living adjacent to sludge-applied fields. The decision unit (DU) size was the same as the exposure unit, which was determined to be 80 acres, the land area that could be worked by a farmer in a day.
The MDNR Hazardous Waste Program (HWP) took the lead role in the investigation and enrolled field sampling help from MDNR Environmental Service Program (ESP). MDNR also worked with the MDHSS and the EPA. Through a cooperative agreement with the EPA, MDNR conducted a pre-CERCLIS site screening to determine the sites eligibility for entry onto CERCLIS.
The planning team for this project included various project managers in both the HWP and the ESP, and chemists in the ESP Chemical Analysis Section (CAS). Additional planning team members included EPA Region 7, the Agency for Toxic Substances and Disease Registry (ATSDR), EPA Technology Innovation Field Services Division (TIFSD), Tetra Tech and Applied Speciation Laboratories.
As recommended in Triad project planning, the laboratory was involved with the project before investigation samples were sent.
The project team divided a farm field into 80-acre DUs. Each DU was then gridded to have 80 1-acre subunits called sampling units (SUs). The number of SUs to sample was determined statistically based on CSM information and DMA data. Statistical analysis of the DMA’s Cr6 data predicted that 3 SUs per DU were sufficient. To avoid under-estimating how many SUs needed to be sampled, SU locations in the farm field were chosen to maximize the concentration differences between the SUs, which would cause the statistics to call for more SUs to be sampled. In this way, a "worst case" scenario was created to ensure sufficient sampling was performed during the first mobilization. Sampling within an SU was performed using ICS. The number of increments to use per SU was determined during the DMA by evaluating the degree of short-scale heterogeneity in total chromium concentrations. Since concentration variability was relatively low, 10 increments per SU was statistically found to be more than sufficient. The 10 increments were combined into a single SU-ICS sample which was then processed by drying, disaggregation, and sieving.
After sample processing, samples were analyzed with an XRF instrument to provide rapid turnaround of total chromium results. Statistical assessment of the total chromium data was conducted in real-time using pre-programmed spreadsheets and VSP to verify that the number of increments initially collected within each SU (10) and the initial number of SUs sampled per DU (3) were sufficient to statistically manage the variability of Cr6 across the fields. Low variability in total chromium concentrations predicted low variability in Cr6 concentrations. Statistical analysis via VSP confirmed that the initial sample numbers were more than sufficient to meet statistical data quality objectives (DQOs). Once the statistical representativeness of the SU samples was verified, equal portions from the SU incremental samples (SU-Iss) from a DU were combined to form the DU incremental sample (DU-IS) which represented the average Cr6 concentration across the 80-acre DU. A representative subsample of the DU-IS was then sent for laboratory Cr6 analysis. The ISM design is illustrated graphically in Figure 1.
Mean background Chromium VI values were also generated in nearby fields in the same manner and compared to mean DU Chromium VI values to determine if a release of Chromium VI can be documented. DU Chromium VI values were also compared with site-specific screening levels to evaluate the health threat to farmers living in or near sludge impacted fields.
Real-time measurement technologies included the use of an FP-XRF instrument in combination with VSP software and field generated statistics. The farm-field pilot study established a relationship between total Chromium variability and Chromium VI variability from laboratory results, verifying the use of an FP-XRF analyzer, which only measures total Chromium. An "Action Level" value of 1,720 mg/kg was calculated for total Chromium using the site-specific Chromium VI screening level of 86 mg/kg and adjusting it using a conservative value for the ratio of total to Chromium VI from the farm fields pilot study (Table 1).
TQRS not prepared
As discussed previously, real-time data management tools were critical to the success of the investigation and included the following [Assessing Past Use of Tannery Waste Sludge as Farm Field Fertilizer in Northwest Missouri]:
To update this profile, contact Cheryl T. Johnson at Johnson.Cheryl@epa.gov or (703) 603-9045.