Old tanneries became a concern in Wilmington, Delaware with the discovery of arsenic (a known carcinogen) concentrations in surface soil above industrial limits (RBCs). The excavation of decomposed animal hides also raised fears of anthrax exposure. Further investigation identified 53 tannery "complexes" including 168 tanneries on historic atlases, which affected more than 350 current landowners. This presentation will review the use of a single GIS and Web-enabled database to manage and present in real-time on-going historical research, field survey, prioritization, investigation, and remediation of these locations to multiple audiences inside and outside of the public domain.
Wilmington, Delaware was formerly the second-largest center of tannery operations on the East Coast. Due to the hazards associated with arsenic and other metals related to tannery operations Tetra Tech, Inc. (Tetra Tech) was contracted by the Delaware Department of Natural Resources and Environmental Control (DNREC) Site Investigation and Remediation Branch (SIRB) to perform a citywide screening for tannery-related sites in Wilmington, Delaware. The goal of this investigation was to locate areas in and near Wilmington, Delaware that were used for tannery operations during the 19th and 20th Centuries, with particular attention to locations where the historic use of arsenic and other metals was likely. Tannery operations were identified from various sources and prioritized by their possible impact to human health or the environment. This prioritization scheme was used to schedule sampling efforts at each site in order to characterize the extent of contamination.
A large number of properties are associated with the tannery sites, and numerous governmental, public and private organizations have an interest in knowing the current location and status of these locations. To make the results of the investigation and sampling available, and to promote the rapid, simultaneous distribution of accurate and current knowledge about the tanneries without frequent correspondences, a project website was constructed. A GIS/database format was already in use and with minor adjustments it was able to serve as a vehicle for driving the website. One database was used for storing site-related data, producing reports, serving the Website, and input/tracking of sample analysis results. The use of a single central database for all aspects of the project provided consistent and accurate information to a large number of stakeholders in a timely and efficient manner.
The sheer volume of information to be collected during this investigation demanded a well-designed database model. This database must track multiple concurrent tasks and serve information to produce reports, maps and web pages. Additionally, the data structure must be portable for sharing with other government agencies that may or may not have much in the way of GIS infrastructure. Due to these limitations a Microsoft Access database was constructed which could be loosely coupled to ArcGIS shapefiles of site outlines and sampling areas. This structure allowed the same database to also be used behind an active server page website providing real-time investigation status to the public and other interested parties. Separate tables within the database were used to store data on historical findings, details of the field investigation, site photos, land parcel data and sampling results. Information collection was organized and will be presented here around four key objectives:
From historical documents and information gathered from residents of the City of Wilmington, Tetra Tech developed a listing of the city's tanneries. Historical references included the Baist Atlas of 1901, the G.W. Beers Atlas of 1868, the G.W. Hopkins C.E. Survey of 1876, the Franklin Survey Co. Atlas of 1936 and the Sanborn Map Company Fire Insurance Maps of 1884, 1901, 1927, 1948 and 1967-1984. Atlases were obtained from the Delaware Historical Society, University of Delaware Morris Library and the City of Wilmington Planning Department. While a substantial list, it must be understood that research of this type can provide good information on many tanneries, but cannot guarantee that all tanneries have been identified. Multiple overlapping tannery data were sorted and grouped into related tannery operations according to ownership and geographic similarities.
The largest part of this effort was thoroughly examining every available map within these historical atlases. A conservative approach governed the investigation, intended to provide a more complete list of tannery operations rather than to exclude possible tanneries. Once a tannery or tannery-related industry was identified, the map was digitally scanned (where possible) or photocopied for future analysis.
Tanneries typically overlapped on multiple information sources. To identify unique tannery operations a tracking number was assigned to catalogue each scan by year and tannery. This tracking number, the image name, historical tannery name, and historical map source were entered into a Microsoft Access database table. On the scanned image the tannery location was identified, the image was georeferenced, and the outline of tannery-related structures and storage areas as illustrated on the document was entered into a GIS layer. At the conclusion of the historical investigation, nine digital data layers (one for each data source) were merged into a single shapefile for a total of 168 tannery outlines. Based on their locations these outlines were grouped and identified with codes (some previously established by DNREC to track environmental sites). This methodology produced a total of 53 clusters subsequently referred to as tannery "sites."
DNREC investigation/remediation were ongoing or complete at some tanneries. Investigation and remediation coincident with tannery locations were identified by DNREC SIRB and provided to Tetra Tech as outlines on land parcel boundaries. These outlines were added to the existing GIS as a separate digital data layer and classified as either "Completed" or "In Progress."
A "windshield survey" to each tannery site identified multiple characteristics that would influence site prioritization, including:
On subsequent visits original observations were verified and photographed to capture the general appearance of the area and document exposed soil. Tetra Tech visited every tannery site on at least four separate occasions.
All information collected during this phase was entered into a separate Microsoft Access database table that recorded and tracked field investigation progress. Another table was developed to store and manage digital site photos including photo name, location and date. Site maps were produced for each location, stored as images and likewise catalogued using an Access table. This structure allowed photos to be available for reporting, the website and in ArcGIS through hotlinks.
Because of the large number of sites identified (53) a prioritization scheme was developed based on selected criteria that were considered potentially relevant human exposure factors. These factors were selected to provide objective criteria that could be used by DNREC for assigning resource priorities and identifying any necessary action. Prioritization was provided only as one of several possible approaches for recommending a "next step" for the sites identified, and was not intended to represent the views of any agency of government; rather it represented Tetra Tech's professional judgment supported by site observations. This prioritization scheme represented a qualitative comparison of observed site characteristics, and was not intended to be a substitute for a quantitative assessment of risk.
Four separate factors guided our prioritization scheme based on potential exposure:
These factors were considered in the order presented, with the exception of known DNREC sites that were automatically deferred to the lowest priority because they are the subjects of ongoing or completed investigation/remediation efforts. Based on the presence of these factors, sites were classified into one of four priority levels:
Although this prioritization system follows a strict methodology, it was subjective in nature and was not in any way a static ranking. New information was and is added as it becomes available and is incorporated into the prioritization, particularly from sampling results that could establish a quantifiable justification for more in-depth investigations.
With prioritization completed, the surface soil investigation began at Initial Action Sites. This action built upon the existing database. Site locations were queried and the underlying tax parcels from New Castle County, Delaware selected. Tax parcel information was linked to ownership data obtained from 1st Real Estate Solutions. This information was used in mail-merged documents to contact residents about the potential hazards and notify them of sampling and meeting dates.
Following resident and owner notification sampling proceeded. Hand augers were used to collect surface soil samples from each exposure area identified in prior field visits. Sampling locations were biased towards areas of historical tannery use (especially beam houses where hides were hung to dry) and current areas of exposed soil or active use (gardens or playgrounds). Sampling locations were collected via field measurement or GPS where appropriate. At these urban sites the use of GPS was difficult as buildings and other structures often obscured the signal. Field measurements however, coupled with quarter foot resolution imagery allowed sample locations to be tracked accurately. Rather than produce another shapefile, the coordinates of each sampling location along with sampling time, sampler and sample characteristics were recorded in a database table.
Samples were sent off for inorganic analysis at a DNREC laboratory with a smaller subset of duplicates sent for confirmatory analysis at a private lab. This data was recorded in separate tables to allow comparison between the two sources. Analyte values may be displayed in ArcGIS by linking to the sampling table. The format of this part of the database closely follows the EPA's Electronic Data Deliverable Format for ease of transfer among agencies.
A final database table was added with the State of Delaware's regulatory limits for inorganics. This table was used to compare against laboratory results and highlight exceedances as new data was added. The website, reports and maps all utilize this table.
While the database was used to support and manage this investigation, the advantages of this centralized approach were most apparent in reporting and analysis. The first major deliverable for this project was a 500+ page report of findings reviewing the historical, field visit and site prioritization. The sheer size of this report made reproduction prohibitive. While the database approach allowed one to retrieve individual reports for each tannery, the number of stakeholders and rapidly changing status of this investigation demanded a better method of dissemination.
To facilitate public access to sampling progress Tetra Tech developed a project website for DNREC. Active Server Pages display all the results of investigation, from scans of historical documents to current sampling results. This format allows visitors to go directly to their neighborhood or browse the report. Visitors may zoom to a location, search by address or navigate by site number. While ArcIMS could have been used, a system of image maps proved just as efficient with improved download speed and a substantially reduced cost. Address searches are performed against the New Castle County Tax Database in a many to many relationship with a table of properties coincident with tannery sites. This linking table was developed using a coincident query in ArcGIS.
As the investigation continues, the project website has been continually refined. Ongoing soil sampling is currently entering into an expanded phase with detailed investigations taking place at a number of locations where concentrations were identified above background levels in the initial surface soil investigations. Remediation techniques are also under development and a select group of soil samples will be sent off for anthrax analysis. The database structure for this project is proving its functionality, tracking new and sometimes unexpected data. The website and project GIS have both become a living documents as the working structure is daily fed new information to display.
The central database approach provides a means by which duplicity is avoided, multiple goals may be accomplished with a minimum of effort and maximum functionality is afforded. While the details of this project may be unique to a small audience, the general lessons learned are applicable to many studies:
If a project website will be included, spend some time to consider your site visitors' backgrounds. Consider the following in your project site design:
Baist, G.W., Property Atlas of the City of Wilmington, Delaware. 1901.
Beers, J.G., Atlas of the State of Delaware. 1868.
Delaware Department of Transporation, New Castle County Road Centerlines. 2001.
Franklin Survey Co., Atlas of Wilmington, New Castle County, Delaware, Volume One. 1936.
Hopkins, G.M., City Atlas of Wilmington, Delaware. 1876.
Chadwyck-Healey, Sanborn Fire Insurance Maps, 1884 (Delaware microform). Originally pub. by: Sanborn Map Co., New York) Teaneck, NJ: 1982.
Chadwyck-Healey, Sanborn Fire Insurance Maps, 1901 (Delaware microform). Originally pub. by: Sanborn Map Co., New York) Teaneck, NJ: 1982.
Chadwyck-Healey, Sanborn Fire Insurance Maps, 1927 (Delaware microform). Originally pub. by: Sanborn Map Co., New York) Teaneck, NJ: 1982.
Chadwyck-Healey, Sanborn Fire Insurance Maps, 1948 (Delaware microform). Originally pub. by: Sanborn Map Co., New York) Teaneck, NJ: 1982.
New Castle County, Delaware, Cadastral Database and Maps. 2001.
Sanborn Map Co., Sanborn Fire Insurance Maps, 1967-84. Sanborn Map Co., New York: 1984.
United States Department of the Interior, Geological Survey, 7.5 Minute Series Digital Line Graph for Wilmington North, Del.-N.J., scale 1:24000. 1987.
United States Department of the Interior, Geological Survey, 7.5 Minute Series Digital Line Graph for Wilmington North, Del.-N.J., scale 1:24000. 1993.
Wilmington, Delaware, Aerial Grayscale Orthophotography. 1997.