| GISdevelopment.net ---> GITA 1999 ---> Engineering and Design Applications |
GIS-Based Data Management for Environmental Investigations
Lisa G. Taylor and Mary E. House
Woodard & Cur-ranInc
41 Hutchins Drive, Portland, ME 04102
Lisa G. Taylor and Mary E. House
Woodard & Cur-ranInc
41 Hutchins Drive, Portland, ME 04102
Introduction
Investigation of industrial properties and hazardous waste sites is a time-consuming and sometimes costly process, in part because the field effort can take many weeks or months to complete. This field effort, which generally includes documentation of hydrogeologic data and collection of environmental samples for laboratory analysis, is typically performed in stages in order to allow time for data processing and interpretation. As such, it often takes several phases, or mobilizations, of field work to fully characterize environmental conditions and estimate contaminant extent.
In an effort to streamline information management and eliminate mukiple field events, we used geographic information system (GIS) software to evaluate environmental conditions at a 10-acre, uncontrolled hazardous waste site in northern New England. Due to an accelerated schedule, we had to assess the volume of contaminated material at this site in less than eight weeks, which ruled out multiple field events. Use of GIS software coupled with electronic data transfer and automated uploading, enabled real-time decision making so that we could complete the investigation in one mobilization.
The purpose of this paper is to demonstrate the potential time and cost savings associated with the use of GIS software during the interpretation of environmental conditions at industrial and hazardous waste sites. We describe the information management process and system software, including electronic data transfer, uploading, real-time interpretation, and reporting, and evaluate the use of this technology with a case study. Lastly, we provide a summary of the costs incurred on this project and compare these with the estimated costs expended using non-GIS-related methods.
Environmental Data Management
Characterization of environmental conditions at industrial and hazardous waste sites requires evaluation of potentially voluminous quantities of detailed geologic, hydrologic, and chemical data. In order to complete a field investigation in one phase, we need to evaluate these data in real-time so that the field crew can modifi exploratory work and delineate contaminant extent, with fewer samples, in one mobilization. Use of GIS-based software augmented with standardized impotiexport routines allows us to manage these data so that they can be retrieved easily and presented effectively, all within an acceptable time period and budget.
The Data Management Process
Environmental data management encompasses the collection and documentation of analytical and physical information, sample tracking, production and maintenance of a comprehensive, electronic database, and manipulation of this database for characterization of site conditions (Figure 1). The process begins with the selection or design of a suitable, electronic basemap, development of templates and sorting parameters for populating the database, and compilation of available, historic data. Once field work has commenced, the project scientists begin tracking samples and documenting conditions at the site. Physical information (e.g., soil descriptions, depth to groundwater, contacts between geologic units, etc.) is typically uploaded to the database as it is received from the field; sample tracking data (e.g., sample ID, collection date, sampling interval, and required tests), are electronically compiled (preprocessed) with analytical data for construction of import templates and database uploading. Once the data have been validated and/or reviewed, real-time analysis of analytical and hydrogeologic data can begin.
Figure 1. Environmentaldata managementflow chart
System Software
The software used for this study consisted of an integrated collection of off-the-shelf packages and customized macros designed to reduce the costs associated with the production, maintenance, and manipulation of complex, environmental databases. The core software, GISKeyTM,was developed by GIS\Solutions of Concord, CA for the management, interpretation, and visualization of geologic, hydrologic, and chemical data. This windows-based, desktop application uses a Visual FoxPro@database (with a Visual Basic@interface), integrated with AutoCAD@ using Visual Basic”, Lisp, and ACS/ARX (versions of C++). Information in the database is linked to AutoCAD@ drawings via exploration locations, such as soil borings, wells, and sampling points.
GIS\KeyT”includes automated query tools and graphics applications for reporting (graphing, mapping, tabulating, and exporting); identification of suspect data (e.g., blank contamination, action level exceedences); and drawing management, as well as user-editable libraries of chemical names, CAS numbers, analytical test methods, and standards. Because of its open architecture and use of commercial off-the-shelf products, GIS\KeyT”can be used with other graphics platforms (i.e. ArcView@),attached to other database repositories, or queried directly using numerous third-party database tools (USEPA, 1994).
Microso~ Excel was used for sample tracking, manual entry of some hydrogeologic information, and preprocessing of analytical data. A customized macro written in Microso~ Visual Basic for Applications by Planet One GIS Software of Portland, ME was used to compile sample tracking information with analytical data.
Data Transfer
Prior to the initiation of field work, we negotiated electronic deliverable formats with the selected laboratory so that they could automatically download requisite analytical data directly from their laboratory information management system to GIS\Keym’s import templates (reformatted Microsok Excel spreadsheets). This up-front coordination resulted in a data transfer system designed to eliminate manual data entry and associated transcription errors, reduce preprocessing time, and produce electronic files ready for importing.
At the same time, we began manually populating the sample tracking spreadsheet with proposed sampling information, including sample ID’s, collection intervals, and required test methods. This ensures that all requisite samples are collected and permits us to print sample container labels, thus eliminating double entry of data, unnecessary time in the field handwriting labels, and the potential for the laboratory to run the wrong test or mis-identify samples based on illegible container labels.
The analytical data were e-mailed to Woodard & Curran Inc (W&C) and electronically compiled with sample tracking information (the compiler macro also confirms that the laboratory has processed all samples). The data were then uploaded to the project database using GIS\KeyT”WinBuild, the import routine provided with the software. This routine permits input of a large volume of data at any one time without manual entry and verifies that there are no missing or inaccurate fields that would otherwise render the database useless. Data which fail this step are tagged with exception or warning codes so that they can be easily addressed. This saves significant time (we don’t inspect import templates for errors GIS\KeyT”WinBuild will identi~) and provides added assurance that data in the database will be predictably retrievable.
Physical data were manually entered into Microsoft@Excel spreadsheets and then appended to the project database. Use of electronic files in the field assures the crew that mandatory fields are completed and provides data management personnel in the office with legible information.
Data Interpretation and Reporting
Standard reporting routines provided with the software were used for data interpretation. These include automated queries for posting data to project basemaps, and drawing management tools for creating cross sections, graphs, boring logs, and contour maps. Because these routines take only minutes to run, we can evaluate our data from many perspectives in a short time-frame, and present this information in a meaningful way. Once interpretation has been completed, the figures can be inserted in the final report without additional drafting.
Case Study
W&C was retained to evaluate the nature and distribution of site-related contaminants in lagoons and overburden soils at the subject site. Although the data generated during this study were expected to supplement historic information for development of remedial alternatives, the primary goal was to estimate the volume of contaminated materials so that cost-based decisions could be made before proceeding with the feasibility study. This meant that the volume of contaminated materials at the site had to be ascertained within eight weeks so that the client could meet a December 11 deadline imposed by outside sources.
Most field investigations require more than one trip, or mobilization, to the site. During the initial mobilization, samples are collected and analyzed, and the results are evaluated over a period of several weeks to determine if the nature and extent of contamination have been established. Typically, more samples are required, necessitating additional trips to the site. Because data processing and interpretation seemingly drive the speed at which confirmatory field work can progress, and multiple mobilizations were not possible, we concluded that this aspect of the investigation had to be conducted in real-time in order to meet the December 11 deadline. Based on this, we opted to map site data using GIS software that included automated import and data verification routines. The case study is described with respect to real-time analysis in the following subsections.
Site Backmound
The site is located in a former borrow pit which was excavated for sand and gravel in the mid-1 940’s. Between the early 1950’s and 1980’s, the property was used for recycling waste oil. The oil was processed in AST Area 1 and releases during handling appear to have flowed from here into a series of interconnected ponds at Lagoons 1 through 4 (see Figure 2 for site features). Tank bottom sludges were deposited in the Sludge Spreading Area south of AST Area 1.
In 1978, additional soils were excavated from the lower gravel pit east of the ponds. About the same time, a section of the Sludge Spreading Area was partially filled, effectively dividing the original, interconnected ponds into four discrete lagoons (contaminated materials in the Sludge Spreading Area typically underlie approximately 2.5 feet of relatively clean soil). Since the lagoons were bermed, there have been at least two breaches at Lagoon 4 resulting in a release of oily sludge to the lower gravel pit.
In August 1997, the regulatory agency responsible for site oversight completed a preliminary remedial investigation of the property. The study identified approximately 0.3 feet (ft) of sludge at the bottom of Lagoons 1 and 2, and up to 3 fi of sludge underlain by 5 to 15 ft of contaminated soils at Lagoons 3 and 4, but did not delineate overall extent. These materials were found to contain volatile (VOC) and semi-volatile (SVOC) organic compounds, polychlorinated biphenyls (PCBS), and lead above applicable guidelines.
Real-Time Analysis
Field work was conducted at the site from November 10 through 25, 1997. The program included soil borings, excavation of test pits, and collection of sediment and surface water samples from the lagoons. Certain “primary” samples were obtained first and submitted to Katahdin Analytical Services of Westbrook, ME for chemical analysis of lead and PCBS on a 48-hr. turnaround. These tests were expected to provide preliminary information on the extent of contaminated materials and govern the placement or analysis of additional “secondary” samples. Many of these primary samples were also tested for VOCS, SVOCS, TCLP (toxicity characteristic leaching procedure) parameters, and/or additional inorganic analytes for waste characterization purposes, depending on the detected concentrations of lead and PCBS. Turnaround time on these tests was about two weeks.
Twenty three primary samples were collected from test pits between November 10 and 12, at depths ranging from 1.2 to 7.8 ft below ground surface (bgs). The samples were shipped to the laboratory the day they were obtained, and analyzed within two working days while the field crew completed other tasks. The results were e-mailed to us on November 13, 14, and 17, and uploaded to the project database on the dates they were received.
Figure 2. Site map showing estimated contaminant extent after primary sampling.
Figure 2 shows the estimated horizontal extent of lead and/or PCB-contaminated soil or sludge from 2.5 to 6.0 ft bgs based on the results of the primary samples and visual observation of stained materials. This figure was constructed by querying the database, posting the analytical results on the project basemap, contouring the concentrations of lead and total PCBS above applicable residential guidelines, and then overlaying the resulting contours. Similar figures were also constructed for depths less than 2.5 ft bgs and greater than 6.0 ft bgs.
Based on these data, 12 new test pits were located, excavated and sampled on November 18 and 19, one day after receiving the last of the primary sampling results from the laboratory. The samples from these new pits, and some of those collected earlier (on hold at the laboratory pending test results of the primary samples) were then analyzed and contoured with the primary results to better define contaminant extent west of the Sludge Spreading Area, east of Lagoon 4 at the lower gravel pit, and west of Lagoons 1 and 3 (Figure 3). These secondary samples enabled us to complete our analysis of contaminant extent and no further samples were required.
The entire field investigation took 15 days. In addition to the lead and PCB maps, we contoured VOC and SVOC data from the waste characterization tests to confirm that the extent of contamination was well documented. These maps, and those produced to assess contaminant concentrations relative to adult worker guidelines, as well as various cross sections, sludge thickness maps, and bedrock surface plans (22 figures in all), were then used to calculate the volume of soil and/or sludge that could require remediation or containment at the site. These were presented to the client on December 11, 1997.
Figure 3. Site map showing estimated contaminant extent afler secondary sampling. Borings were drilled for primary purposes, but the results from these samples were not received until afier the secondary samples were tested.
Time and cost savings
Figure 4 presents the schedule required to complete each project task vs. the time that could have been required if the investigation had not been conducted in real-timel. Real-time interpretation saved time by eliminating multiple field events (thus decreasing our time in the field), which in turn eliminated ( 1) time spent writing multiple work plans (one for each field phase); (2) collection of too many samples (which can occur when data interpretation is not completed concurrent with field work), and (3) excess hours at the report-writing stage detailing the many activities involved in multiple field events. We also saved time by using automated routines to produce presentation and report-ready graphics that required little or no redrafting. As a result, this project realized a time savings of about 8 to 10 weeks.
Saving time saves money, and the shortened schedule produced significant cost savings. Figure 5 provides a breakdown of the costs expended on this project vs. those anticipated using a non GIS-based approach to site investigations. As expected, the largest savings were realized from the field investigation. Because this was conducted in real-time, we saved money by (1) mobilizing equipment to the site only once, (2) writing one work plan, (3) collecting the minimum number of samples required to establish contaminant extent (this is possible only if the sampling equipment remains onsite during data interpretation), and (4) eliminating unnecessary analytical tests. The remaining cost savings resulted from the construction of report-ready graphics during data interpretation which did not require redrafting during report preparation. Overall, we saved an estimated 30 percent on the expected costs for this investigation.
Figure 4. The project schedule required real-time analysis of environmental data so that we could complete this investigation in one mobilization and report on the volume of contaminated materials within eight weeks. A typical investigation normally consists of multiple field phases, which could have increased the project schedule by 8 to 10 weeks.
Conclusion
Real-time interpretation of environmental data significantly reduces field-related costs by facilitating conduct of site investigations in one mobilization. These savings are achieved through the use of integrated software and GIS spatial analysis which allows us to manage and interpret large volumes of complex, environmental data in a short time-frame. Used in association with pre-formatted, electronic laboratory deliverables and 48-hr. turnaround on analytical results, we can modi~ ongoing exploratory work based on a continuously-evolving knowledge of site conditions and estimate contaminant extent without the expense of remobilization. Additional savings are realized by creating report-ready graphics during data interpretation.
The real-time investigation of an uncontrolled hazardous waste site in northern New England permitted the field crew to complete contaminant delineation in about half the time required for a typical, multi-phased investigation. Corollary cost savings were an estimated 30 percent. Although the primary driver behind these savings was an accelerated schedule, it is anticipated that the benefits of real-time data management can be applied to most environmental investigations.
Figure 5. Costs expended for this project were approximately one-third less than costs for similar projects which used multiple phases of field investigation. Cost savings were realized by completing the investigation in one mobilization (due to real-time analysis of environmental data) and using GIS-based software to complete report-quality figures during data management and interpretation.
Acknowledgements
Wewould liketothank the following W&Cemployees forpeer review andgraphics suppoti: Chigako Wilson, Ted Taylor, Karl Kasper, and Jason House. A special thanks to Anne Tischbein, project manager for the case study, who offered valuable insights into case study activities.
References
Investigation of industrial properties and hazardous waste sites is a time-consuming and sometimes costly process, in part because the field effort can take many weeks or months to complete. This field effort, which generally includes documentation of hydrogeologic data and collection of environmental samples for laboratory analysis, is typically performed in stages in order to allow time for data processing and interpretation. As such, it often takes several phases, or mobilizations, of field work to fully characterize environmental conditions and estimate contaminant extent.
In an effort to streamline information management and eliminate mukiple field events, we used geographic information system (GIS) software to evaluate environmental conditions at a 10-acre, uncontrolled hazardous waste site in northern New England. Due to an accelerated schedule, we had to assess the volume of contaminated material at this site in less than eight weeks, which ruled out multiple field events. Use of GIS software coupled with electronic data transfer and automated uploading, enabled real-time decision making so that we could complete the investigation in one mobilization.
The purpose of this paper is to demonstrate the potential time and cost savings associated with the use of GIS software during the interpretation of environmental conditions at industrial and hazardous waste sites. We describe the information management process and system software, including electronic data transfer, uploading, real-time interpretation, and reporting, and evaluate the use of this technology with a case study. Lastly, we provide a summary of the costs incurred on this project and compare these with the estimated costs expended using non-GIS-related methods.
Environmental Data Management
Characterization of environmental conditions at industrial and hazardous waste sites requires evaluation of potentially voluminous quantities of detailed geologic, hydrologic, and chemical data. In order to complete a field investigation in one phase, we need to evaluate these data in real-time so that the field crew can modifi exploratory work and delineate contaminant extent, with fewer samples, in one mobilization. Use of GIS-based software augmented with standardized impotiexport routines allows us to manage these data so that they can be retrieved easily and presented effectively, all within an acceptable time period and budget.
The Data Management Process
Environmental data management encompasses the collection and documentation of analytical and physical information, sample tracking, production and maintenance of a comprehensive, electronic database, and manipulation of this database for characterization of site conditions (Figure 1). The process begins with the selection or design of a suitable, electronic basemap, development of templates and sorting parameters for populating the database, and compilation of available, historic data. Once field work has commenced, the project scientists begin tracking samples and documenting conditions at the site. Physical information (e.g., soil descriptions, depth to groundwater, contacts between geologic units, etc.) is typically uploaded to the database as it is received from the field; sample tracking data (e.g., sample ID, collection date, sampling interval, and required tests), are electronically compiled (preprocessed) with analytical data for construction of import templates and database uploading. Once the data have been validated and/or reviewed, real-time analysis of analytical and hydrogeologic data can begin.
Figure 1. Environmentaldata managementflow chart
System Software
The software used for this study consisted of an integrated collection of off-the-shelf packages and customized macros designed to reduce the costs associated with the production, maintenance, and manipulation of complex, environmental databases. The core software, GISKeyTM,was developed by GIS\Solutions of Concord, CA for the management, interpretation, and visualization of geologic, hydrologic, and chemical data. This windows-based, desktop application uses a Visual FoxPro@database (with a Visual Basic@interface), integrated with AutoCAD@ using Visual Basic”, Lisp, and ACS/ARX (versions of C++). Information in the database is linked to AutoCAD@ drawings via exploration locations, such as soil borings, wells, and sampling points.
GIS\KeyT”includes automated query tools and graphics applications for reporting (graphing, mapping, tabulating, and exporting); identification of suspect data (e.g., blank contamination, action level exceedences); and drawing management, as well as user-editable libraries of chemical names, CAS numbers, analytical test methods, and standards. Because of its open architecture and use of commercial off-the-shelf products, GIS\KeyT”can be used with other graphics platforms (i.e. ArcView@),attached to other database repositories, or queried directly using numerous third-party database tools (USEPA, 1994).
Microso~ Excel was used for sample tracking, manual entry of some hydrogeologic information, and preprocessing of analytical data. A customized macro written in Microso~ Visual Basic for Applications by Planet One GIS Software of Portland, ME was used to compile sample tracking information with analytical data.
Data Transfer
Prior to the initiation of field work, we negotiated electronic deliverable formats with the selected laboratory so that they could automatically download requisite analytical data directly from their laboratory information management system to GIS\Keym’s import templates (reformatted Microsok Excel spreadsheets). This up-front coordination resulted in a data transfer system designed to eliminate manual data entry and associated transcription errors, reduce preprocessing time, and produce electronic files ready for importing.
At the same time, we began manually populating the sample tracking spreadsheet with proposed sampling information, including sample ID’s, collection intervals, and required test methods. This ensures that all requisite samples are collected and permits us to print sample container labels, thus eliminating double entry of data, unnecessary time in the field handwriting labels, and the potential for the laboratory to run the wrong test or mis-identify samples based on illegible container labels.
The analytical data were e-mailed to Woodard & Curran Inc (W&C) and electronically compiled with sample tracking information (the compiler macro also confirms that the laboratory has processed all samples). The data were then uploaded to the project database using GIS\KeyT”WinBuild, the import routine provided with the software. This routine permits input of a large volume of data at any one time without manual entry and verifies that there are no missing or inaccurate fields that would otherwise render the database useless. Data which fail this step are tagged with exception or warning codes so that they can be easily addressed. This saves significant time (we don’t inspect import templates for errors GIS\KeyT”WinBuild will identi~) and provides added assurance that data in the database will be predictably retrievable.
Physical data were manually entered into Microsoft@Excel spreadsheets and then appended to the project database. Use of electronic files in the field assures the crew that mandatory fields are completed and provides data management personnel in the office with legible information.
Data Interpretation and Reporting
Standard reporting routines provided with the software were used for data interpretation. These include automated queries for posting data to project basemaps, and drawing management tools for creating cross sections, graphs, boring logs, and contour maps. Because these routines take only minutes to run, we can evaluate our data from many perspectives in a short time-frame, and present this information in a meaningful way. Once interpretation has been completed, the figures can be inserted in the final report without additional drafting.
Case Study
W&C was retained to evaluate the nature and distribution of site-related contaminants in lagoons and overburden soils at the subject site. Although the data generated during this study were expected to supplement historic information for development of remedial alternatives, the primary goal was to estimate the volume of contaminated materials so that cost-based decisions could be made before proceeding with the feasibility study. This meant that the volume of contaminated materials at the site had to be ascertained within eight weeks so that the client could meet a December 11 deadline imposed by outside sources.
Most field investigations require more than one trip, or mobilization, to the site. During the initial mobilization, samples are collected and analyzed, and the results are evaluated over a period of several weeks to determine if the nature and extent of contamination have been established. Typically, more samples are required, necessitating additional trips to the site. Because data processing and interpretation seemingly drive the speed at which confirmatory field work can progress, and multiple mobilizations were not possible, we concluded that this aspect of the investigation had to be conducted in real-time in order to meet the December 11 deadline. Based on this, we opted to map site data using GIS software that included automated import and data verification routines. The case study is described with respect to real-time analysis in the following subsections.
Site Backmound
The site is located in a former borrow pit which was excavated for sand and gravel in the mid-1 940’s. Between the early 1950’s and 1980’s, the property was used for recycling waste oil. The oil was processed in AST Area 1 and releases during handling appear to have flowed from here into a series of interconnected ponds at Lagoons 1 through 4 (see Figure 2 for site features). Tank bottom sludges were deposited in the Sludge Spreading Area south of AST Area 1.
In 1978, additional soils were excavated from the lower gravel pit east of the ponds. About the same time, a section of the Sludge Spreading Area was partially filled, effectively dividing the original, interconnected ponds into four discrete lagoons (contaminated materials in the Sludge Spreading Area typically underlie approximately 2.5 feet of relatively clean soil). Since the lagoons were bermed, there have been at least two breaches at Lagoon 4 resulting in a release of oily sludge to the lower gravel pit.
In August 1997, the regulatory agency responsible for site oversight completed a preliminary remedial investigation of the property. The study identified approximately 0.3 feet (ft) of sludge at the bottom of Lagoons 1 and 2, and up to 3 fi of sludge underlain by 5 to 15 ft of contaminated soils at Lagoons 3 and 4, but did not delineate overall extent. These materials were found to contain volatile (VOC) and semi-volatile (SVOC) organic compounds, polychlorinated biphenyls (PCBS), and lead above applicable guidelines.
Real-Time Analysis
Field work was conducted at the site from November 10 through 25, 1997. The program included soil borings, excavation of test pits, and collection of sediment and surface water samples from the lagoons. Certain “primary” samples were obtained first and submitted to Katahdin Analytical Services of Westbrook, ME for chemical analysis of lead and PCBS on a 48-hr. turnaround. These tests were expected to provide preliminary information on the extent of contaminated materials and govern the placement or analysis of additional “secondary” samples. Many of these primary samples were also tested for VOCS, SVOCS, TCLP (toxicity characteristic leaching procedure) parameters, and/or additional inorganic analytes for waste characterization purposes, depending on the detected concentrations of lead and PCBS. Turnaround time on these tests was about two weeks.
Twenty three primary samples were collected from test pits between November 10 and 12, at depths ranging from 1.2 to 7.8 ft below ground surface (bgs). The samples were shipped to the laboratory the day they were obtained, and analyzed within two working days while the field crew completed other tasks. The results were e-mailed to us on November 13, 14, and 17, and uploaded to the project database on the dates they were received.
Figure 2. Site map showing estimated contaminant extent after primary sampling.
Figure 2 shows the estimated horizontal extent of lead and/or PCB-contaminated soil or sludge from 2.5 to 6.0 ft bgs based on the results of the primary samples and visual observation of stained materials. This figure was constructed by querying the database, posting the analytical results on the project basemap, contouring the concentrations of lead and total PCBS above applicable residential guidelines, and then overlaying the resulting contours. Similar figures were also constructed for depths less than 2.5 ft bgs and greater than 6.0 ft bgs.
Based on these data, 12 new test pits were located, excavated and sampled on November 18 and 19, one day after receiving the last of the primary sampling results from the laboratory. The samples from these new pits, and some of those collected earlier (on hold at the laboratory pending test results of the primary samples) were then analyzed and contoured with the primary results to better define contaminant extent west of the Sludge Spreading Area, east of Lagoon 4 at the lower gravel pit, and west of Lagoons 1 and 3 (Figure 3). These secondary samples enabled us to complete our analysis of contaminant extent and no further samples were required.
The entire field investigation took 15 days. In addition to the lead and PCB maps, we contoured VOC and SVOC data from the waste characterization tests to confirm that the extent of contamination was well documented. These maps, and those produced to assess contaminant concentrations relative to adult worker guidelines, as well as various cross sections, sludge thickness maps, and bedrock surface plans (22 figures in all), were then used to calculate the volume of soil and/or sludge that could require remediation or containment at the site. These were presented to the client on December 11, 1997.
Figure 3. Site map showing estimated contaminant extent afler secondary sampling. Borings were drilled for primary purposes, but the results from these samples were not received until afier the secondary samples were tested.
Time and cost savings
Figure 4 presents the schedule required to complete each project task vs. the time that could have been required if the investigation had not been conducted in real-timel. Real-time interpretation saved time by eliminating multiple field events (thus decreasing our time in the field), which in turn eliminated ( 1) time spent writing multiple work plans (one for each field phase); (2) collection of too many samples (which can occur when data interpretation is not completed concurrent with field work), and (3) excess hours at the report-writing stage detailing the many activities involved in multiple field events. We also saved time by using automated routines to produce presentation and report-ready graphics that required little or no redrafting. As a result, this project realized a time savings of about 8 to 10 weeks.
Saving time saves money, and the shortened schedule produced significant cost savings. Figure 5 provides a breakdown of the costs expended on this project vs. those anticipated using a non GIS-based approach to site investigations. As expected, the largest savings were realized from the field investigation. Because this was conducted in real-time, we saved money by (1) mobilizing equipment to the site only once, (2) writing one work plan, (3) collecting the minimum number of samples required to establish contaminant extent (this is possible only if the sampling equipment remains onsite during data interpretation), and (4) eliminating unnecessary analytical tests. The remaining cost savings resulted from the construction of report-ready graphics during data interpretation which did not require redrafting during report preparation. Overall, we saved an estimated 30 percent on the expected costs for this investigation.
Figure 4. The project schedule required real-time analysis of environmental data so that we could complete this investigation in one mobilization and report on the volume of contaminated materials within eight weeks. A typical investigation normally consists of multiple field phases, which could have increased the project schedule by 8 to 10 weeks.
Conclusion
Real-time interpretation of environmental data significantly reduces field-related costs by facilitating conduct of site investigations in one mobilization. These savings are achieved through the use of integrated software and GIS spatial analysis which allows us to manage and interpret large volumes of complex, environmental data in a short time-frame. Used in association with pre-formatted, electronic laboratory deliverables and 48-hr. turnaround on analytical results, we can modi~ ongoing exploratory work based on a continuously-evolving knowledge of site conditions and estimate contaminant extent without the expense of remobilization. Additional savings are realized by creating report-ready graphics during data interpretation.
The real-time investigation of an uncontrolled hazardous waste site in northern New England permitted the field crew to complete contaminant delineation in about half the time required for a typical, multi-phased investigation. Corollary cost savings were an estimated 30 percent. Although the primary driver behind these savings was an accelerated schedule, it is anticipated that the benefits of real-time data management can be applied to most environmental investigations.
Figure 5. Costs expended for this project were approximately one-third less than costs for similar projects which used multiple phases of field investigation. Cost savings were realized by completing the investigation in one mobilization (due to real-time analysis of environmental data) and using GIS-based software to complete report-quality figures during data management and interpretation.
Acknowledgements
Wewould liketothank the following W&Cemployees forpeer review andgraphics suppoti: Chigako Wilson, Ted Taylor, Karl Kasper, and Jason House. A special thanks to Anne Tischbein, project manager for the case study, who offered valuable insights into case study activities.
References
- USEPA, 1994, GIS\Key Environmental Data Management System, Innovative Technology Evaluation Report: EPA/540/R-94-505, 131 p.