Section II: Findings and Data Analysis
Description of Labor skills Needed for Cleanup: A Study of 18 Sites

Section III: Demand For Jobs Over Time
A. Projections Of The Overall Number Of Workers Needed To Complete Mandated Hazmat Cleanup Mandated Hazmat
    Cleanup

Bibliography

Appendices
Appendix I: Methodology
Appendix II: Sites in the Data Base
Appendix III: Tables Referred to in Text
Appendix IV: Tables Not in Text but Available on Request
Appendix V Forms (not available in html)
Appendix VI: Acronym List
Appendix VII: Glossary

Table of Tables and Charts

Table 1: Percent of Gross Pay by Predominant Category at 17 Sites
Table 2: Labor Mix at 11 NPL Sites, Based on Gross Pay from Certified Payrolls
Table 3: Average Hourly Earnings By Predominant Category of Worker at 11 NPL Sites
Table 4: Labor Costs as a Percentage of Total Remediation Costs at 5 NPL Site
Table 5: Labor Market Projections for Remedial Action and O&M Jobs. Based on University of Tennessee "Best Estimate"
Table 6: Labor Market Projections for Remedial Action and O&M Jobs. Based on Department of Energy "High Cost Alternative"
Table 7: Cost of Hazardous Waste Cleanup: University of Tennessee Estimates (Billion of Dollars)
Table 8: Cost of Hazardous Waste Cleanup: Department of Energy Estimates (Billions of 1992 Dollars)
Table 9: Range of Estimates For Cost of Superfund Cleanup
Table 10: Breakdown of NPL Costs by Function
Table 11: Projected NPL Construction Labor Job Years On. 13 Estimates Of NPL Cleanup Costs (Over 30 years)
Table 12: Projected NPL Remedial Action Job Years by Craft
Table 13: Operations and Maintenance Costs Compared to Capital Costs at 11 NPL Sites
Table 14: Percent Increase of Employment Episodes. Planning by Overall Jobs vs. Peak Month Jobs
Table 15: Billions of Dollars That Have Been and Will Be Spent on Various Categories of Hazardous Waste Cleanup 1990-2010
Table 16: Number of Expected Remedial Action Jobs Generated by Various Categories of Hazardous Waste Cleanup 1990-2010
Table 17: Number of Expected Remedial Action Jobs by Craft, By Years for Hazardous Waste Cleanup 1990-2010

Chart 1: Crafts as a Percent of Gross Pay
Chart 2: Lipari Landfill: Community Income and Site Wages Compared (not currently available in html)
Chart 3: Dollars That Have Been and Will be Spent on Various Categories of Hazardous Waste Cleanup 1990-2010
Chart 4: Number of Expected Remedial Action Jobs Generated by Various Categories of Hazardous Waste Cleanup 1990-2010
Chart 5: Numbers of Expected Remedial Action Jobs by Craft, By Years for Hazardous Waste Cleanup 1990-2010

Cleaning up the nation's hazardous waste sites is an enormous undertaking, requiring the efforts of millions of workers and hundreds of billions of dollars. On-site remedial action alone, 1990-2010, will utilize three million job years, or 4.5 billion hours, of labor. Operations and maintenance work will require another one billion labor hours.

Who are the workers that the nation is relying on to do the cleanup work? They are construction workers, industrial workers, transportation workers, and emergency responders. Many industrial and construction workers have been trained in hazardous materials health and safety and will bring the skills of their trades to hazardous waste site cleanup. Additional individuals are gaining skills both in hazardous materials handling and in the specific industrial and construction skills necessary to cap a landfill, monitor pump and treat activities, or remove underground storage tanks. Hundreds of thousands of emergency personnel will respond to spills, leaks, fires, explosions, and other hazardous materials incidents. Hazardous waste cleanup is labor intensive.

A better understanding of the labor market for hazardous waste workers can make the planning of environmental cleanup more efficient, and safety and health training more targeted and timely. The Worker Education and Training Program at the National Institute of Environmental Health Sciences (NIEHS) and the Environmental Protection Agency (EPA), Office of Solid Waste and Emergency Response contracted with Ruth Ruttenberg & Associates, Inc. (RRA) to study the labor market associated with hazardous waste cleanup work. The actual experience at a number of hazardous waste sites across the nation have been recorded from certified payroll information. RRA has developed an extensive data base from these payrolls and has analyzed a broad range of other labor market data to create a profile of the current cleanup work force and to make projections of future work. The focus of this study is the construction, industrial, transportation, and emergency response workers necessary to clean up hazardous waste sites.

OVERVIEW OF HAZARDOUS WASTE CLEANUP

For more than 40 years the disposal of hazardous waste at landfills, industrial plants, military bases, and other locations across the country has contaminated many thousands of sites and nearby communities. Environmental and public health risks identified at these hazardous waste sites include contaminated air, direct contact with hazardous waste, contaminated drinking water, ecological damage, fire and explosion hazards, exposure through the food chain, and contaminated groundwater, soil, and surface water.¹

How Many Sites Are There?

Hundreds of thousands of sites must be cleaned up. There are approximately 1285 sites on the National Priorities List and tens of thousands of sites regulated by RCRA. There are thousands of sites under the jurisdiction of the Department of Energy and the Department of Defense. More than 295,000 underground storage tanks require closure or removal and there are as many as 425,000 state and private sites.

The Cleanup Process is Complex

The cleanup of a hazardous waste site many take many years. The process is arduous and complex. Preparation for the actual remediation of soil or groundwater may occur only after years of investigations and negotiations. Much time and money has been spent on site investigations and feasibility studies, costing out each remedy and listing the advantages and disadvantages of each treatment plan.

Upon first notification of an incident or potentially hazardous site, the appropriate regulatory body performs a preliminary assessment (PA) to determine whether action is necessary. If the PA indicates an emergency requiring immediate or short-term action to reduce risk to the public, a removal action is conducted to stabilize or clean up the site. After the removal action, if a hazard remains, a site inspection (SI) is conducted to determine if a site warrants scoring under the Hazard Ranking System (HRS) -- a system that scores sites on the potential effects from contamination on human health and the environment. Those sites which score 28.5 and higher are proposed for the National Priorities List (NPL) of EPA.

If a site is placed on the NPL, an in-depth planning and investigation phase -- called remedial investigation (RI)/feasibility study (FS) -- takes place. The results of the RI/FS and the rationale for selecting a remedy are required by EPA and are documented in a Record of Decision (ROD). In some instances, several RI/FSs and RODs are needed for different operable units -- portions of the site -- which require separate cleanup actions. RODs specify the technology type deemed to be the appropriate remedy for a site, and after many years of experience, and scientific and engineering evaluations of performance data on technology use, EPA has developed guidelines for the most appropriate remedies for treating specific types of sites. These preferred treatments for common categories of sites are referred to as presumptive remedies. Presumptive remedies are not available for all contaminant types but pilots underway should help make remedy selection and, thereby, the writing of RODs increasingly more efficient. RODs also list the alternatives that were considered, and the pros and cons of each selected alternative.

Using the ROD, detailed engineering specifications for the selected cleanup alternative are developed. These designs are then used to solicit bids for remedial action (RA). Operations and maintenance (O&M) activities, if necessary, begin at the conclusion of remedial action. O&M activities include groundwater monitoring, periodic site inspections, and other activities designed to ensure continued effectiveness of the remedial actions.

Sites which do not rank high enough on the HRS still need to be cleaned up. These are typically addressed through state programs, which go through similar steps.

As cleanup proceeds, many different crafts are needed on-site for the complex array of activities which occur over the course of cleanup. Usually a range of tasks are performed simultaneously. For example, at Lipari Landfill work that was going on in March 1990 included excavation, foundations, and concrete work for buildings and tanks. These tasks lasted through October 1990, while in June of that year work began on the plant plumbing and lasted until mid-September. Also in July, work began on the heating, ventilation, and air conditioning duct work for the plant, and in August outside tanks were erected. Well drilling activities occurred throughout this time. In addition to the main contractor, there was major work done by the Army Corps of Engineers and at least twenty-two individual subcontractors. More than a dozen labor crafts were involved.

Another example of the multiple on-going activities at a complex site is the cleanup at Paducah Gaseous Diffusion Plant. In the spring of 1994 site decontamination pads were built, access control fences were installed, drilling and environmental restoration staging activities were underway as well as plume well installation and clamshell building construction. In just two and a half months, no fewer than 15 different crafts and at least thirteen contractors performed work.

A prime or general contractor is responsible for getting the cleanup done. In addition there may be a company which acts as project coordinator. The general contractor hires subcontractors, and those subcontractors may in turn, hire others. Pinpointing exactly what company is working on what part of what task and when -- and for which company workers are doing actual remediation work -- may be challenging.

The focus of this study is on-site containment and on-site treatment. Further research could provide estimates for both ex-situ activities for containment and treatment and also employment associated with the building of treatment equipment and even pre-fab buildings off-site which later come on-site for cleanup purposes. The on-site craft labor employment studied here is only one portion of the job opportunities associated with hazardous waste cleanup work.

Economic Impact.

Hazardous waste cleanup not only improves the environment, but it also introduces new economic opportunities to a community. Hazardous waste cleanup is a growing sector of the economy. In the first three years of the decade alone, waste management activities grew by more than 40 percent.² A draft economic impact study done for the Environmental Protection Agency in 1994 found that for every $1 of Superfund expenditure, $3.10 of goods and services were generated.³ Over the period FY81-FY92, this multiplier meant that $7.6 billion of Superfund procurement had total direct and indirect economic impact of $23.5 billion.⁴ The overall impact that cleanup of hazardous waste will have on the U.S. economy is difficult to project since not all sites have been identified; many site assessments have not been completed; some sites will require more or different cleanup than originally anticipated; and the longer cleanup takes, the more expensive it will likely become.⁵ But if NPL cleanup is estimated to cost at least $100 billion, the direct and indirect generation of demand for goods and services associated with this aspect of Superfund, is well over $300 billion. Understanding the mechanism of how environmental restoration expenditures may positively impact community, as well as national, economic base and job generation provides important insights for coordination of economic and environmental policies.

One recent draft EPA study⁶ determined that Superfund activities had created 32 jobs per million dollars spent, either directly or indirectly, for craft labor and supervisory activities, by federal contractors and manufacturers of cleanup equipment. This RRA study focuses only on those physically cleaning up the site and conservatively estimates 4.4 direct craft labor jobs associated with each million dollars spent. The portion of the cleanup dollar that goes to actual earth moving, capping, chemical treatment, pumping of contaminated ground water, etc. is likely to be more labor-intensive -- and labor-intensive with the kind of manufacturing and construction jobs that add economic base and economic growth to an area. As hazardous waste sites are remediated, land can often be returned to productive use, thus further fueling the engine of economic growth, strengthening new businesses and labor markets alike.

Is There Such a Job Category As Hazardous Waste Worker?

As hazardous waste cleanup expands and intensifies, millions of individuals will spend a portion of their working lives in the hazardous waste cleanup process. Cleanup requires a significant amount of skilled labor, but there is no formal job category that can be identified as hazardous waste worker. The Standard Industrial Classification (SIC) system of the U.S. Department of Commerce does not provide an analytic category for hazardous waste worker. The SIC system does include SIC 4953, Hazardous Waste Worker, for those responsible for the destruction of hazardous refuse. SIC 4212 includes transportation workers who haul hazardous waste. But these SIC codes are far from inclusive. There are many other individuals who work at hazardous waste sites. And if one asks these individuals to describe their job, they are more likely to say, "I am a carpenter;" "I am a heavy equipment operator;" or "I am a pipe fitter" than "I am a hazardous waste worker." These individuals may spend most of their working lives on hazardous waste sites. More likely, they will work a number of jobs within their trade -- many of them unrelated to hazardous waste remediation. Ongoing research will help to identify the degree to which those receiving hazardous materials training may increasingly focus their career on hazardous waste cleanup and the degree to which hazardous waste cleanup is but one type of job that individuals have over the course of their work life.

For the purposes of this study, anyone working at a hazardous waste site, especially if in an area requiring HAZMAT training, has been included in the data base. There are several jobs which one may not have originally characterized as HAZMAT jobs. For instance, people hired to fence a site are considered HAZMAT workers for the purposes of this study. They are listed on the certified payrolls of cleanup contractors and subcontractors, and they usually require HAZWOPER training. The same is true for those who build the pump and treat systems, roads, or do landscaping after the waste is removed. The hazardous waste site remediation process has had a direct impact on producing these jobs. One does not need to wear personal protective clothing to be performing HAZMAT work. Also included, but often harder to fully identify, are emergency responders who often work off-site and transportation workers who move hazardous waste off-site for treatment or containment.

INTRODUCTION TO THIS STUDY AND THE DATA UPON WHICH IT IS BASED

The objective of this study is to describe and project the number and types of jobs needed to clean up the nations hazardous waste sites. This was accomplished using detailed data from the existing U.S. hazardous waste remediation labor force. The focus of data collection was federal cleanups, because they provide detailed Davis Bacon certified payrolls which catalog job categories, hours, and pay. These federal sites include those under the responsibility of EPA, DOE, and DOD.

This study has pioneered the use of very detailed certified payroll records to document real experience in expenditures of time and money at hazardous waste cleanup sites. The data document differences and similarities of cleanup sites across the country -- in labor mix, earnings, and scheduling of work. From such a "real experience" data base, more accurate labor market profiles can be established and more solidly based projections can be made. (See Appendix I for detailed Methodology.)

Data obtained from certified payrolls included standard and overtime hours worked, hourly pay, and job category. For most of the sites studied, certified payrolls were available for only a portion of the remediation work at that site. More often the payrolls for the data base represented discrete tasks such as constructing a decontamination pad, building a pump and treat system, or drilling wells.

An objective in data collection was to obtain data from sites where there had been task completions, not necessarily site completions. Some data is from the late 1980s, but most is from the mid 1990s. Time lines for task completions at specific sites were developed as part of the study process and are available on request. All numbers used in the text emanate from the RRA data base unless otherwise footnoted.

This study is based on the collection of more than 80,000 Davis-Bacon⁷ payroll records from 17 sites, daily labor logs from one site⁸, other labor market data from 12 sites, data from EPA's Removal Cost Management System (RCMS) for eight federal removal sites, the analysis of data on more than 20,000 state sites with data collected by EPA in conjunction with ASTSWMO, and creation of data bases and analysis from five emergency response data systems (three in Arizona, one in California, and one in New Jersey), as well as extensive review and analysis of existing studies on environmental remediation work across the United States. The detailed data upon which this study is based provides important information to those who do job planning and/or job training associated with the cleanup of hazardous waste. Data were obtained from cleanup contractors, EPA area offices, DOE field offices, and the Army Corps of Engineers. The certified payroll data base is one which can be expanded over time to remain current and also to more comprehensively cover the nation's cleanup activities. The data are also a source for future research, a rich resource of actual experience from which a wide variety of analyses and projections can flow.

Protection of Privacy

Of primary importance in the data collection process was the privacy for individual workers whose names appeared on certified payrolls. No personal identifies were used in the RRA data base -- only labor category. Copies of the forms presented to contractors and government officials outlining and assuring a privacy protocol are in Appendix V. (Also in Appendix V is a sample certified payroll form.)

Sites for Which Certified Payroll Data Were Obtained

Efforts were made to obtain certified payroll data documentation at dozens of sites across the country. The eighteen sites for which detailed data were obtained represent a broad range of site ownership, geographical location, causes of contamination, types of contamination, and types of remedial action. (See Appendix II for a brief description of each of these sites.) Had full access to data been available, more private and RCRA sites would be in the sample.

The eighteen sites which are the focus of the primary RRA data base represent sites of private industry, the U.S. Department of Defense, and the U.S. Department of Energy⁹. The sites are in thirteen states and twelve are on EPA's National Priorities List (NPL).¹⁰ Activities which contributed to the contamination at these sites include landfills, surface impoundments, wellfields, leaking containers, asbestos hazards, radiological tailings, and waste oil. Four sites are landfills; four are uranium mill tailing remediation sites; and the other ten include a mix of activities. A broad range of remedial action categories are represented in the data base: ¹¹ five sites used institutional controls, including monitoring, access restriction, and/or alternate water supplies; four sites used on-site containment as a remedy; at least eight sites used water collection/treatment/discharge; two sites used oil/sediment removal, low intensity treatment, and site restoration; four sites used soil/sediment removal, high intensity treatment, ash disposal, and site restoration; at least three sites used in-situ treatment; nine sites used soil/sediment removal and landfilling; and one site used water collection/discharge to existing facility. (See Tables B & C.)

The sites studied are both small and large -- ranging from $500,000 at Cherokee County to more than $14 million at BROS. Studied were specific tasks in the cleanup processes at each of the 18 sites. Only at the Durango UMTRA site did RRA obtain 100 percent of the certified payrolls for 100 percent of the cleanup work, from beginning of remedial action through completion. Because some of the sites are particularly large and complex and because many of them still have cleanup tasks pending, payroll data are for only specific tasks over specific months. RRA obtained data for as much as 75 percent of payroll, as at Lipari, but in other instances, the data obtained by RRA represented only 15 percent of total estimated payroll at a site.

The eighteen sites which make up the primary RRA data base are:

• California Sacramento Army Depot
• Colorado Durango Uranium Mill Tailings Action (UMTRA) Site
• Colorado Grand Junction UMTRA Site
• Colorado Rifle UMTRA Site
• Florida Hollingsworth Solderless Terminal Company
• Kansas Cherokee County, Galena Subsite
• Kentucky Paducah Gaseous Diffusion Plant
• Louisiana Bayou Bonfouca
• Missouri Kem-Pest Laboratories
• New Jersey Bog Creek Farm
• New Jersey Bridgeport Rental & Oil Services (BROS)
• New Jersey Lipari Landfill
• New Jersey Lone Pine Landfill
• New Mexico Shiprock UMTRA Site
• Ohio New Lyme Landfill
• Pennsylvania Moyer Landfill
• Tennessee K-25 Gaseous Diffusion Plant
• Washington South Tacoma Channel, Well 12A

Sites for Which Other Than Certified Payroll Data Were Obtained

When it was not possible to obtain certified payrolls for a site, RRA obtained other labor market data, that contractors would make available. For the twelve sites where "other" data were obtained, that data included information on job descriptions, the composition of job crews, hours worked, and/or equipment used.

Ten of these sites were in California: Concord Naval Weapons Station, Embarcadero, Hunter's Point Naval Shipyard, Intel, Mather Air Force Base, McClellen Air Force Base, Pillar Point Air Force Base, Raytheon, Richmond Harbor, and a battery plant. In Missouri data came from Weldon Springs, in New Jersey from Ciba Geigy, in Ohio from Reactive Metals Inc., and in Tennessee from X-10 and Y-12 at Oak Ridge.

Sites of the EPA Removal Cost Management System

There are more than 1500 completed site cleanups in the data bases of the EPA Removal Cost Management System (RCMS). RRA was successful in obtaining data on eight sites from four EPA regions in seven states. The data available from this system include predominant category of worker, average hourly earnings, and hours worked. Removals differ from remedial action sites in that they are typically smaller with a higher level of contamination and require cleanup on an "emergency" basis.

The eight sites for which data were obtained are: Turner Seed Company in Iowa, Superior Polishing in Michigan, Bannister Road Drum in Missouri, Carolina Creosote II in North Carolina, Anderson Residential Lead in South Carolina, Chemet in Tennessee, and Bernard Neal and Martinsburg Drum Dump in West Virginia.

Sites of the EPA/ASTSWMO State and Territorial Data Base

Cleanup of the largest number of hazardous waste sites is under the jurisdiction of the states and territories. A 1994 EPA-sponsored study of these sites produced a data base of over 20,000 sites from 39 states and two territories. The sites varied considerably, with cleanup costs ranging from $1000 to $7 million. Some sites were remedial actions and some were removals. They covered a wide range of contaminants and a wide range of cleanup remedies. The data base is useful in characterizing the similarities and contrasts among sites and the ways that sites are remediated by the various states.

Data on Emergency Response Associated with Hazardous Materials Incidents

Analysis of the labor market associated with emergency response poses challenges different from the labor market for other hazardous waste workers. Few emergency responders have jobs dedicated to a single site or even to hazardous materials response alone. On-site emergency response personnel often work full time at other tasks. Off-site, usually public sector, emergency response personnel are not assigned to a specific site, and are not usually part of a dedicated HAZMAT team.

The emergency response focus of this study is to characterize, to the extent possible, how many incidents require emergency response, what the nature of these incidents is, where they occur, how long they take to resolve, what emergency response personnel are called, and the nature of injuries. These data provide important perspective on the labor needs and training needs of emergency response organizations.

Among the issues reviewed, is a frequent concern of emergency response personnel -- that many of the incidents and injuries associated with a given hazardous waste site may not occur on the site itself, but rather along a transportation corridor as materials are being transported for off-site containment or treatment. The data collected for this report concur with this concern.

Data to study emergency response to hazardous materials incidents came from Arizona, California, and New Jersey. Automated data bases on emergency response associated with hazardous materials incidents are available in only a few states. RRA was successful in obtaining automated data from the state of Arizona, which maintains three separate data bases: the Arizona Hazardous Materials Incidents Reports (AHMIR), the Emergency Response Notification System (ERNS) data sheets, and reports from the Fire Departments. The state of California, through its Office of Emergency Services on Hazardous Materials Incidents, compiles its data into an annual report, which RRA reviewed and analyzed. Six months of individual, hard copy hazardous materials incident reports of the New Jersey Department of Environmental Protection were obtained and entered into a separate data base for analysis.

Literature Review

This labor market study began with an extensive literature review. As the study proceeded, literally hundreds of documents, commentaries, and other studies were reviewed -- both for relevant data and for strengthened perspective.

ORGANIZATION OF THIS REPORT

Section I is an introduction to the study -- describing its objectives, scope, the data bases upon which it is based, and the methods used for data collection, followed by an overview of the labor market for hazardous waste cleanup.

Section II reports on data findings and data analysis. This section is divided into three parts. First are overall findings. Second are more detailed findings and analysis on projections for the largest categories of hazardous waste cleanup; i.e., NPL and federal removal sites under the jurisdiction of Superfund, RCRA, the Department of Energy, the Department of Defense, the Underground Storage Tank program and state and private sites. Finally, other issues are presented, including the impact of remedy choice on the labor market, emergency response labor market needs, issues relating to jobs and environmental justice.

Section III interprets the findings, and moves on to projections for future hazardous materials jobs, from 1990 until 2010, as well as future needs for health and safety training.

The activities of workers at hazardous waste sites are documented in Section II. Findings from the certified payroll data base -- findings of job categories, hours worked, wages paid, residence of workers, and equipment used are presented and analyzed. These and other data are used in Section III to project future demand for hazardous waste workers and associated emergency responders.

DESCRIPTION OF LABOR SKILLS NEEDED FOR CLEANUP: A STUDY OF 18 SITES

Categories of Labor

Construction labor, industrial labor (often O&M workers), transportation workers, and emergency responders, dominate hands-on remedial action and hazardous waste activities. Hazardous waste sites differ in labor skills needed. Nearly all sites rely heavily on the work of equipment operators, laborers, and truck drivers. Other skilled labor often in demand, sometimes in significant demand, are asbestos workers, boilermakers, bricklayers, carpenters, cement workers, chemical workers, drillers, electricians, emergency responders, iron workers, machine operators, mechanics, painters, plumbers and pipe fitters, rail workers, roofers, stationary engineers, and sheet metal workers. Certified payroll records, the foundation of the RRA data base, documented the use of these and other labor categories. The RRA data base covered over two million work hours and more than $40 million of payroll. Three crafts -- operators, laborers, and truck drivers -- accounted for more than 60 percent of the payroll. Eight crafts -- operators, laborers, truck drivers, carpenters, electricians, mechanics, plumbers and pipe fitters, and iron workers -- accounted for 80 percent of the payroll. (See Table 1 and Chart 1.) Data collected included dollars spent for craft labor activities at a site (see Table D), the number of standard and overtime hours devoted to cleanup, hourly rates earned -- and sometimes the levels of personal protection used in the field, equipment used, and the towns where workers lived.

From site to site, the relative use of crafts is quite variable. (See Table 1 and E.) At a few sites plumbers and pipe fitters had a significant presence -- most notably at Lipari Landfill in New Jersey, where they accounted for 27 percent of the gross pay for the time period studied. Carpenters, who were non-existent in the certified payrolls at some sites, made up nearly 20 percent of gross pay at Paducah. Iron workers, while accounting for nearly 15 percent of the gross payroll at Paducah for the months studied, were only represented in eight of the eighteen key sites. Laborers, while present at all sites, varied in their intensity of use at a site from 3 percent at the Durango and Grand Junction UMTRA sites in Colorado to 56 percent of gross payroll at Kem-Pest in Missouri and 70 percent of gross payroll at Hollingsworth in Florida. Operators were present at all sites, and their use ranged from 2 percent of gross pay at Takoma in Washington State to 64 percent at Shiprock. Electricians earned 20 percent of the gross pay at the South Tacoma Channel Site, and mechanics earned nearly 10 percent of gross pay at Grand Junction.

Key:
CASAAD - Sacramento Army Depot
Hollingsworth - Hollingsworth Solderless Terminal Company
BROS - Bridgeport Rental & Oil Services

Source: Ruth Ruttenberg & Associates, Inc. Data Base.

In-plant industrial workers performing on-going cleanup tasks, while not always captured in the certified payroll data, are major participants in the cleanup process. Industrial workers are employed by chemical companies, by facilities of the nuclear weapons complex, by military bases, or by companies maintaining a RCRA site. Industrial workers are often responsible for operations and maintenance. They maintain operations, monitor, and are responsible for the day to day running of plants and facilities. Major O&M activities include maintenance, reporting, sampling and analysis, project management, pump and treat, and oversight engineering. Also a critical part of the hazardous waste labor force are emergency responders -- some on site; most employed in neighboring communities and along transportation corridors.

Definition of a labor category is not always simple and because definitions may vary from site to site, there is some reporting variability from site to site. Laborers might operate equipment, do carpentry work, or drive vehicles. At some sites these workers would be identified as laborers; at others they would be identified as operators, carpenters, and drivers. For example, a company noted for its work crew models,¹² establishes some crew categories that involve only laborers but which also include hydraulic cranes. One might assume that hydraulic crane operators would be operators and not laborers. Another example would be a crew category of laborers, with the equipment used being highway trucks, when one might assume that these jobs would be assigned to truck drivers and not to laborers. Common work practice also makes job definition more complex. Over a given work day or work week, one individual may work a composite of jobs -- doing the work of a laborer for a few hours, an equipment operator for a few hours, and a truck driver or a mechanic or a carpenter for a few hours. In addition the work of truck drivers sometimes is recorded in inconsistent or incomplete ways.¹³

Wages

Data analysis from certified payrolls at the 17 sites for which payroll data were available indicated average hourly earnings of $18.40 -- including overtime, shift differential, hazard pay, etc. (See Table F.) Site averages ranged from $9.49 at Hollingsworth in Florida, where the payroll was not covered by Davis-Bacon and labor was primarily industrial and non-union, to $32.50 at Bog Creek in New Jersey where wages were higher, perhaps due to significant amounts of overtime and call pay. The other 15 sites had average hourly earnings ranging from an overall average of $12.05 at Cherokee County in Kansas to $23.85 at Sacramento Army Depot in California.

Use of Overtime

Overtime ranged from 2 percent of total hours at Shiprock in New Mexico to 24 percent of total hours at Hollingsworth in Florida. The use of overtime at the other 15 sites in the RRA data base ranged from 5 percent of total hours at Durango in Colorado to 23 percent of total hours at Kem-Pest in Missouri. The average use of overtime across the 17 sites was 16 percent of total hours worked.

Some of the data, at Lipari for example, suggest that there was a learning curve in planning for labor, and overtime hours fell over time. At other sites it appeared that overtime was affected by peak demands for certain crafts; i.e.., carpenters or electricians or pipe fitters or iron workers, whose individual jobs were for intense but rather short durations. The certified payrolls themselves, however, do not indicate the reason for overtime.

Where Workers Live

Local employment is the predominant source of cleanup labor. Residential data were available for twelve sites¹⁴ in the RRA data base. At most sites for which residential data were collected, the majority of payroll went to workers living within 11-25 miles of the site. Typically, 50 percent to 80 percent of the payroll went to workers who lived within 25 miles of the site. Sometimes the majority of workers lived extremely close to the site or within community boundaries. At Lipari, 65 percent of the payroll went to workers who lived within 10 miles of the site. At Shiprock, located on the Navajo Nation, over 98 percent of the payroll went to those living on the Navajo Nation; nearly 50 percent of the total was earned by those who lived within the town of Shiprock itself. On one occasion, however the majority of the work force lived at a distance. At Bog Creek, nearly 85 percent of the payroll went to workers who lived more than 50 miles from the site.

Equipment Needed For Cleanup

Equipment lists -- some quite brief others; more complete -- were obtained by RRA from five of the sites studied. (Full equipment lists are available on request.) At Lone Pine equipment lists were part of the Daily Labor Log that recorded daily site activity. Data on the type of equipment used, and the number of times and days a given piece of equipment was used, were recorded. Dozers were by far the most used piece of equipment at Lone Pine, accounting for 1530 days, or over 25 percent of total mechanical days. Backhoes --including backhoe excavator, backhoe loader, backhoe with sheer, and backhoe/loader combo -- accounted for nearly 15 percent of the total mechanical days at Lone Pine. Rollers and trucks were 13 percent each; loaders were over 6 percent, and other equipment included chippers, compactors, graders, trenchers, and water tanks.

At Bridgeport Rental & Oil Services (BROS), equipment lists were available on the frequency of use by type of equipment. Several types of backhoes were used at BROS, as well as Lone Pine. Also used at BROS were 5 different types of dozers; 20 types of loaders; 15 types of pumps, forklifts, backhoes; and several types of compressors, cranes, and trackhoes.

Data on equipment from Concord Naval Weapons Station came from an interview with the contractor (ACCI). Eighteen major pieces of equipment and miscellaneous light equipment accounted for 13,120 equipment hours spent in cleanup over a fourteen month period. Equipment needed to cleanup contaminated wetlands included low ground pressure tractors, supersuckers, dozers, excavators, a backhoe, a pressure washer, dump trucks, and pickup trucks.

Equipment lists at Weldon Spring Site Remedial Action Project were available by task. Although the first three tasks listed are the dismantling of buildings, the amount and types of equipment used varied across those three dismantling tasks. Heavy equipment for building removal included staketrucks, roll-off trucks, excavators, cranes, flatbed trucks, shears, pickers, forklifts, and backhoes. Tasks involved with quarry bulk waste removal and temporary storage required waste haul trucks, dump trucks, half-ton trucks, flatbed trucks, low boys, dozers, excavators, tractors with boom, concrete pulverizers, graders, cranes, and excavators.

An equipment list from one week of work was obtained from the Rifle UMTRA site. Many trucks, scrapers, and a few dozers were used for this job, which included the excavation and relocation of contaminated soil. (The Rifle equipment list also provided information on equipment utilization once the equipment was on site -- including truck miles, weather hours, operating hours, repair hours, and idle hours.)

An EPA review of 100 hazardous waste sites¹⁵ found that the 10 most used pieces of equipment (in descending order) -- each used on at least one-third of the 100 sites reviewed -- were:

1. Backhoe/excavator
2. Front-end loader
3. Lowboy
4. Bulldozer
5. Generator
6. Hand tools (shovels, hammers, etc.)
7. Pressure washer/laser
8. Diaphragm pump
9. Air compressor
10. Tractor (OTR)

Discussed below are six major categories of sites and a summary of the labor needs for each. Site categories are: Superfund, RCRA, DOE, DOD, UST, and State/Private sites.

1. Sites Regulated by EPA's Superfund Program

Superfund has two basic types of cleanups: long-term remedial actions and short-term removal actions. To perform a Superfund remedial action, EPA must go through the formal process of placing a site on its National Priorities List. In the Superfund removal program, actions are taken to mitigate immediate and significant threats, such as those stemming from contaminated drinking water or unrestricted access to hazardous waste sites. Removals are generally of a short-term and emergency nature.

A. National Priorities List

Background on NPL Sites

Congress passed the Comprehensive Environmental Response, Compensation, and Liability Act (CERCLA) in 1980 to clean up highly contaminated hazardous waste sites. CERCLA gave EPA the authority to clean up these sites or to require that the parties responsible for the hazardous waste clean them up. CERCLA established a $1.6 billion trust fund, financed primarily by taxes on crude oil and certain chemicals, for EPA to implement and pay for the program. In 1986, the Superfund Amendments and Reauthorization Act (SARA) set new requirements and authorized an $8.5 billion increase in the trust fund. In 1990, Congress reauthorized CERCLA through 1994 and added $5.1 billion to the trust fund authorization. Legislation to reauthorize Superfund is pending before the Congress.

The nation's most hazardous known waste sites are on the National Priorities List (NPL), which had 1285 sites as of April 1995. Cleanup of these Superfund sites (and most other sites as well) is a multi-step process which includes: Site Discovery, Preliminary Assessment (PA), Site Inspection (SI), listing on NPL (if applicable), Remedial Investigation/Feasibility Study (RI/FS), Record of Decision (ROD), Remedial Design (RD), Remedial Action (RA), and deletion from NPL list (if applicable). This study focuses on the labor requirements of the RA Stage.

Findings About NPL Sites

Findings From Certified Payrolls at NPL Sites

The 12 NPL sites which are part of the RRA data base,¹⁶ included a broad array of site types: four landfills, three surface impoundments, two leaking containers, one wellfield, three radiological tailings, one electrical, one waste oil, two manufacturing, one metalworking. (See Table G.) In addition, four of the 13 largest Superfund sites -- representing 7 percent of the active sites FY87-93, but 40 percent of the costs -- are among those included in the RRA data base.¹⁷

Labor Mix. Detailed certified payrolls were analyzed at eleven Superfund sites in eight states, creating a data base of 60,000 records representing more than $31 million of payroll (See Table H) and 1.5 million hours of work. These sites represented a broad array of contaminant types and were a mix of private and federal facilities, ranging from relatively small to extremely large cleanup jobs.¹⁸ Operators earned nearly $10 million, or 31 percent of the total gross pay; followed by laborers, who earned, over $6 million or 19 percent of the gross pay. Other crafts, including plumbers/pipe fitters, carpenters, electricians, truck drivers, iron workers, mechanics, and cement workers, earned between one and six percent of the total gross pay at the eleven NPL sites. (See Table 2.)

As additional sites were added to the RRA data base, both Superfund and Non-Superfund, the labor mix remained relatively stable, despite the variance from one specific site to another. (See Table 1 for a breakdown of percent of gross pay by craft by site).

Wages. Data analysis found average hourly earnings of $20.87 -- including overtime, shift differential, hazard pay, etc. (See Table 3.) Site averages ranged from $9.49 at Hollingsworth in Florida, where the payroll was not covered by Davis-Bacon and labor was primarily industrial and non-union, to $32.50 at Bog Creek in New Jersey where wages were higher, perhaps due to significant amounts of overtime and call pay. The other nine sites had hourly earnings ranging from an overall average of $13.93 at Bonfouca to $23.85 at Sacramento Army Depot. Hourly earnings by craft ranged from an average across sites of $16.79 for laborers to $23.63 for operators.

(1) Laborer category in this instance is primarily industrial labor rather than construction labor.
(2) Average is overall for all crafts, even those not included in this table. Bog Creek average is especially high due to issues of overtime and call pay.

Findings From Other Sites Studied: NPL

In addition to the twelve NPL sites for which certified payrolls were collected, interviews and/or site visits were conducted at other NPL sites as well. Information obtained from interviews with individuals at such sites as Ciba-Geigy, Concord Naval Weapons Station, McClellan Air Force Base, Raytheon, and Weldon Spring, was similar to information found in certified payrolls in terms of craft mix, and wages earned by workers. (Interview summaries and site specific data are available on request.)

Findings About Operations and Maintenance Activities at NPL Sites

What percent is labor of total O&M? O&M costs vary widely from site to site. They may be minimal and they may be greater than the costs of remediation. For six of the thirteen sites in Table J -- Kansas City, Langley, McClellan, two sites at Savannah River, and Twin Cities -- O&M labor as a percent of O&M costs are estimated to be between 12 percent and 40 percent, with a median of 28 percent. Work on pump and treat at Savannah River suggests that approximately one-quarter of O&M labor is for maintenance and three-quarters is for operations. A 1993 DOE study found O&M to be approximately 16 percent of total remediation costs,¹⁹ and to be conservative in projections, this 16 percent estimate is used in this study.

A review of thirteen remediation sites (different from those in the RRA data base.) studied by EPA and DOE (See Table J),²⁰ from a variety of sources, found one year of O&M to cost between 4 percent and 114 percent of total capital costs, with eight or nine of the thirteen sites having total estimated 20 year O&M (undiscounted) costs greater than total capital costs. Thus, in the majority of cases, where O&M is likely to last as long as 30 years, the total O&M cost could be several times greater than the total for capital costs.

B. Federal Removals

Background on Removals

Short-term federal removal actions, often of an emergency nature, are under the jurisdiction of EPA's Superfund Program. As of March 1996 there had been 2,567 completed federal fund-lead removal actions. There have been a total of 3,766 removal actions, which may include more than one removal action per site.²¹ There is however no fixed universe of removals, as sites needing removal actions are continually discovered. EPA estimates that on average there are 220 federal removal actions started each year, with an average of 180 removal actions completed per year.

Findings About Removals Sites

Removal data for eight sites were obtained from EPA's Removal Cost Management System (RCMS). The data came to RRA in automated form with records on labor category and hours worked. The sites were in four of EPA's ten regions. (See Table K.) Each EPA region uses the RCMS system to different degrees; some do not seem to use it often, while others fill out each field of information on a regular basis.

Labor Mix. Based on a detailed study of Removal Cost Management System data for eight removals in seven states in four EPA regions -- which represented an array of removal actions including four time critical and one non-time critical action, those that took two weeks to clean up and those that took eleven months to remediate -- overall labor mix, by number of hours worked, was: laborers 62 percent, equipment operators 35 percent, and drivers 3 percent. Other crafts -- represented in smaller amounts -- included field clerk/typists, chemist/organic, health and safety and truck supervisor, program manager, and secretary. (See Tables L and M.)

Duration of Cleanup. Removal actions, both federal and state, are restricted to either a time limit of one year for completion or a spending cap of $1 million. A new action memo may be written to get a dollar exemption or a time limit extension. A 1994 report²² (which combines state and federal removal data) found that the average duration, for removal-only sites that reported start and completion dates, is 13.89 months (which exceeds the 12 month legally-mandated time limit). On average, completed state and federal removal-only sites have cleanup cost of $3.8 million.

2. Sites Regulated by EPA's RCRA Program

Background on RCRA

EPA, through the Resource Conservation and Recovery Act (RCRA), has developed a basic framework for regulating waste generators, waste transporters, and waste management facilities. RCRA began as an amendment to the Solid Waste Disposal Act in 1965, was passed as a law in 1976, and was amended in 1980 and 1984. The Act provides a regulatory framework for the nation's management of hazardous and solid wastes. Sites covered by RCRA are the largest and most expensive part of the overall environmental remediation budget. RCRA sites include landfills, waste piles, surface impoundments, land treatment units, tanks, tank areas, containment areas, and satellite accumulation areas. RCRA is divided into four programs: underground storage, medical waste, non-hazardous solid waste, and hazardous solid waste. There are a wide variety of wastes located on these sites, the scope of which resembles that of the Superfund universe.

Corrective action site cleanups are funded by site owners or operators. There is no funding program like Superfund to cover sites in the RCRA corrective action universe. The releases from solid waste management units (SWMUs) at treatment, storage and disposal facilities (TSDFs) are included in those remediations covered by the RCRA corrective action program. EPA defines a SWMU as "any discernable unit at which solid wastes have been placed at any time, irrespective of whether the unit was intended for the management of solid or hazardous waste. Such units include areas at a facility at which solid wastes have been routinely and systematically released."²³

While RCRA sites were not the focus of this study, several of the sites studied in this report have ongoing RCRA activities -- Ciba-Geigy, Paducah, and Weldon Spring among them.

Findings About RCRA Sites

Total estimates for the number of TSDFs range from 4700 to 5100,²⁴ with between 1500 and 3500 of the regulated TSDFs requiring corrective action. Beyond this there are an estimated 21,575 large quantity generators; 190,431 small quantity generators; and 2,389 treatment, storage and disposal facilities acting as generators.²⁵

RCRA sites are divided into Solid Waste Management Units. There are approximately 80,000 pre-existing SWMUs at TSDF facilities. Because of the various methods used to treat, store, incinerate, and dispose of many different types of hazardous wastes, operations of these facilities are often complex and costly.

By the end of FY92 corrective actions were underway or completed at 247 facilities, about 3,500 facilities had undergone RCRA facility assessments, and 614 were undergoing RCRA facility investigations. Of these RCRA sites approximately half use off-site disposal remedies and half use innovative treatment technologies. SVE, in-situ bioremediation, and above ground treatment (bioremediation) each make up about a third of the innovative technologies in use.

The workers at RCRA sites are often industrial workers employed by the facility that has a RCRA site to clean up. Collecting data on the labor mix at these cleanup sites is recommended for future research.

3. Department of Energy Environmental Restoration Sites

Background on DOE Sites

For more than 40 years the United States produced nuclear weapons and generated significant amounts of both radioactive and hazardous waste. In 1989, the Office of Environmental Restoration and Waste Management, now called the Office of Environmental Management, at the U.S. Department of Energy, was created to address issues of environmental contamination from DOE production sites.

The Office of Environmental Management has five major tasks, summarized below with an estimated total (75 year) budget of $230 billion:

• Waste Management ($112 billion): To minimize, treat, store, and dispose of DOE waste and to protect worker safety, public health, and the environment. Responsible for hazardous waste, transuranic waste, spent nuclear fuel, sanitary waste, mixed low-level waste, low- level waste, and high-level waste.
• Environmental Restoration ($65 billion): To ensure that risks to the environment and to human health and safety posed by inactive and surplus facilities are either eliminated or reduced to prescribed, acceptable levels. Responsible for landlord activities, assessment, remediation, decommissioning, and surveillance and maintenance.
• Nuclear Material and Facility Stabilization ($22 billion): Responsible for landlord, stabilization, and surveillance and maintenance.
• Technology Development ($12 billion).
• National Program Management and Planning ($19 billion) .

Restoring the environment to safe conditions at nuclear waste sites is achieved through remedial actions and decontamination and decommissioning (D&D). Remedial actions involve cleaning up inactive sites. These actions address contaminated soil and groundwater, as well as sometimes addressing surface-water contamination, tanks, buildings, and structures at active or inactive sites. D&D involves keeping inactive buildings safe until they can be decontaminated, entombed, dismantled and removed, or converted to another use. These structures include reactors, hot cells, processing plants, storage tanks, and other structures. D&D will take place at approximately 500 facilities.

Some of the contaminants at DOE sites are unique to the DOE complex, with mixed waste, containing both radioactive and hazardous waste, constituting a problem at a number of facilities. Other contaminants closely resemble those found on Superfund sites.

Many DOE facilities have both active and inactive sites and therefore must meet the requirements of both RCRA and CERCLA. Thus, DOE and its federal and state regulators often have to coordinate cleanup activities and schedules that stem from the requirements of each Act. DOE's facilities in Oak Ridge, Tennessee are an example of the dual application of RCRA and CERCLA. At Oak Ridge, DOE started corrective action work under RCRA in 1988 on a storage site for solid waste (Waste Area Grouping 6). In 1989, WAG 6 also became subject to CERCLA's requirements when the Oak Ridge site was placed on the NPL. At some DOE facilities, sites regulated under RCRA must be remediated in conjunction with sites addressed under CERCLA. For example, Hanford's B-pond, a disposal site for liquid wastes, is being cleaned up under RCRA, while abandoned trenches that formerly brought wastes to the B-pond are being cleaned up under CERCLA.

Findings About DOE Sites

From the Certified Payrolls

Data from certified payrolls from six DOE sites were obtained by RRA for its labor market data base. These sites were: Durango, Grand Junction, and Rifle in Colorado; Paducah Gaseous Diffusion in Kentucky; Shiprock in New Mexico; and K-25 (Oak Ridge) in Tennessee.

Labor Mix. Labor mix at the six DOE sites in the RRA data base varied significantly, even among the four Uranium Mill Tailings Radiation Control Act (UMTRA) sites. (See Table N.) UMTRA sites, a gaseous diffusion plant, and a former enriched uranium plant pose very different cleanup challenges and tasks, and account for real differences in labor mix.
While most sites predominantly used equipment operators, laborers, and truck drivers, a wide variety of other skilled craft labor, such as roofers, plumbers, and electricians were also used. At Grand Junction, there were a large number of railroad workers employed.

Wages. Average hourly earnings among crafts varied across the DOE sites studied. Plumbers/pipe fitters at Paducah earned an average of $23.00 per hour, while at K-25, they earned an average of $17.77 per hour. Laborers at Paducah earned, on average, $15.32 per hour; at K-25 they earned $12.36; and at Shiprock laborers earned an average of $10.80 per hour. (See Table O.)

Other Sites Studied

Other DOE sites for which interviews were conducted, but certified payrolls were not obtained included Weldon Spring in Missouri and the X-10 and Y-12 Plants on the Oak Ridge Reservation in Tennessee. Workers at Weldon Spring represented a broad array of crafts including carpenters, electricians, laborers, painters/sandblasters, plumbers/pipe fitters, and truck drivers. Tasks included bulk waste removal at the quarry and building removals. Data obtained through interviews at X-10 and Y-12 primarily provided data on industrial workers -- in these instances, chemical workers. Chemical workers at the Oak Ridge facilities earned between $15 and $17 per hour cleaning up mercury, remediating dormant reactors, and performing general D&D activities.

Decontamination and Dismantlement at DOE Sites

Across the DOE complex -- are buildings that, because of their level of contamination -- often radioactive contamination need to be specially treated and then demolished.

Decontamination technologies work at removing surface contamination. They include brushing, scraping, scrubbing, scabbing, vacuuming, pressurized steaming, strippable coating, water jets, shot blasting, grit blasting, pellet blasting, and chemical foams.

The dismantling of a building involves removing yard structures, sealing contaminated structures and working on surface decontamination. The dismantling process also involves the removal of asbestos materials, followed by removal of electrical equipment, piping, water lines, gas lines, tanks, heating, ventilation, etc. The last steps are usually the removal of air filtration apparatus and the removal of the roof, exterior walls, and internal structural members. After above-grade decontamination and dismantlement, foundations, slabs, and pads would be decontaminated or stabilized to minimize further soil contamination. These foundations would then be removed. Dismantled materials would go into interim storage or be transported off-site.

Estimates made at the Fernald Environmental Management Project for its Operable Unit 3 in May 1994²⁶ were that dismantling and decontamination (D&D) would take 16 years and cost $1.076 billion (in 1994 dollars), not including caretaker maintenance. The job would require 6 million person hours of work, not including on-going site O&M.²⁷ Of this total, 9 percent of the work would be to support interim storage activities; the remaining work would be for D&D.

Operating costs for decontaminating include crew costs, chemicals, electrical power and waste disposal. Crew costs include the labor for equipment set-up, decontamination, radwaste treatment, and waste packaging. Typical crews have a range of specific crafts, but generally include a machine or a tool operator and a laborer. Plumbers and pipe fitters are needed for D&D when removal of piping and tanks is needed. Carpenters, electricians, and other crafts are utilized on a less consistent basis for other D&D associated tasks.

Uranium Mill Tailing Remedial Actions Sites (UMTRA)

The Department of Energy in 1978 was directed by Congress to remediate sandlike tailings from mineral processing located at 24 designated sites and at approximately 5000 vicinity properties in ten states and on two Indian Tribal lands. Work was authorized by Public Law 95-604, the Uranium Mill Tailings Radiation Control Act of 1978. The Act directs DOE to provide for stabilization and control of the tailings in a safe and environmentally sound manner. States pay 10 percent of remedial action costs and DOE pays the remaining 90 percent. On the Indian Land sites, the federal government pays 100 percent of the cleanup cost. The tailings were from uranium production of the 1950's, 1960's, and 1970's.²⁸ Work at the UMTRA sites requires the relocation of residual radioactive material to safe disposal sites for long-term stabilization and control. Eleven of the sites were able to dispose of tailings on site. At Shiprock, the contamination was moved only 700 feet from the site, but in the case of Rifle, material was trucked 6 miles and for Grand Junction material was transported by truck and rail 18 miles. Site preparation includes fencing, and the building of temporary roads, drainage ditches, and wastewater retention basins. The next phase of work includes excavation, transportation, and placement and construction of necessary support facilities. Tailings are placed in an underground disposal cell, covered with earthen radon and frost protection layers, drain and filter layers, and a rock erosion protection layer. Surface runoff diversion ditches are usually constructed. Any on-site buildings would likely require decontamination. The site then requires grading and landscaping. DOE has set 1998 as a goal for completing the surface work and 2014 for completing the groundwater components of the project.

The 24 sites are divided into high, medium, and low priority. The four sites for which certified payrolls were obtained are high priority sites: Durango, Grand Junction, and Rifle in Colorado and Shiprock in New Mexico. The four sites compose 33 percent of the volume of contaminated materials at all UMTRA sites (13,000 of 39,000 thousand cubic yards), 18 percent of the acres of contaminated land (700 of 3900 acres), and 88 percent of the vicinity properties (4600 of 5300 properties).²⁹

Findings: UMTRA. Four of the six DOE sites for which certified payrolls were obtained were Uranium Mill Tailings Remedial Action sites. These four sites were: Durango, Grand Junction, and Rifle in Colorado and Shiprock in New Mexico.

UMTRA Labor Mix. Certified payroll data, from four UMTRA sites in Colorado and New Mexico, were reviewed. Actual labor mix varied from site to site. The share of gross pay for operators ranged from 8 percent to 64 percent. Drivers made up 11 percent to 42 percent of payroll. Laborers made up as little as 3 percent and as much as 23 percent of gross pay. (See Table P.) At Grand Junction a large share of the off-site transport was by rail.

While the predominant crafts at UMTRA sites are clearly operators, drivers, and laborers, a number of other crafts were represented as well -- mechanics, carpenters, electricians, cement masons, concrete masons, flaggers, iron workers, pump men, surveyors, trainmen, track washers, welders, working foremen, timekeepers, security personnel, engineers, project managers, superintendents, and clerks. Across the four UMTRA sites, operators accounted for 38 percent of the payroll, drivers 28 percent, and laborers for 17 percent of the payroll. (See Table Q.)

UMTRA Wages. The four UMTRA sites studied had average hourly earnings of $14.94 (including overtime, shift differential, hazard pay, etc). (See Table R.) Site averages ranged from $12.66 at Grand Junction to $15.67 at Durango. The other sites had average hourly rates of $12.82 (Shiprock) and $15.18 (Rifle). Craft averages, across the four sites, ranged from $13.01 for laborers to $20.78 for electricians.

Where UMTRA Workers Live. Significant payroll at three sites was earned by workers residing in the towns of the sites themselves (Data were available for Grand Junction, Rifle, and Shiprock). At least 35 to 47 percent of gross payroll at each of the three sites went to those living in the same cities where the cleanup sites were located. Of all the hazardous waste sites surveyed as part of this study, the UMTRA sites seemed to rely the most heavily on local labor to get the cleanup done. For Rifle, a town of only 4,600 people, nearly $2.5 million in payroll went into the town -- and this, of course, did not include money spent by other workers while in Rifle, as well as money received by businesses for supplies and equipment, etc. In Shiprock, a town of less than 7,700 individuals, where approximately 46 percent of the workforce resided, over half a million dollars in cleanup monies went directly to the community, and many learned new skills as well. Towns affected by a hazardous waste site at least received some of the financial benefits of cleanup through wages for residents and purchases from local merchants.

The definition of local community can vary from site to site. The Shiprock UMTRA site, located on the Navajo Nation, reported over 98 percent of gross pay on its certified payrolls earned by individuals who lived within the Navajo Nation. Though nearly half of that gross pay went to those who lived within the town of Shiprock itself, other Navajos traveled many miles a day to work. Nonetheless, the Navajo Nation, as a community, earned nearly the entire payroll.

4. Department of Defense: Defense Environmental Restoration Program

Background on DOD Sites

Measured by number of contaminated sites, the U.S. military is the nation's largest polluter -- having nearly 12,000 sites at 760 military bases in need of cleanup.³⁰ In addition, the Department of Defense (DOD) is responsible for the cleanup of more than 3,000 sites at 2,200 former domestic bases, many of which contain buried waste. There are 122 DOD sites and 19 former defense properties on the NPL. There are 81 installations with anticipated cleanup bills exceeding $100 million. (Fifty-seven of the 81 DOD facilities whose cleanup is expected to cost more than $100 million are on the NPL.)

DOD has released into the environment the same types of pollution found at privately owned industrial facilities: paints, petroleum products, solvents, and heavy metals. Some facilities have PCB contamination from electrical equipment. Some sites have minor radiation problems, such as radium dials from old aircraft, but the majority of significant radiation problems have been physically transported to Department of Energy sites such as Hanford and the Idaho National Engineering Laboratories.

The most common types of contamination found at DOD sites are landfills, spills, and surface impoundments. Surface discharge areas -- including spills, lagoons and disposal areas -- account for 24 percent of the contamination at active domestic sites. Storage tanks, primarily leaking underground tanks, account for 18 percent of the contamination.³¹ Groundwater is contaminated at 80 percent of the sites.³²

The Deputy Assistant Secretary for Defense has oversight responsibility for the Defense Environmental Restoration Program (DERP), with the Army, Navy, and Air Force responsible for program implementation only. The two major components of DERP are the Installation Restoration Program (IRP), and Other Hazardous Waste Operations (OHW). IRP is responsible for cleanup; OHW is responsible for preventing continued pollution and managing hazardous waste through research, development and demonstration of technologies.³³

The Department of Defense's IRP has four phases similar to the phases involved in other cleanup programs:³⁴

• Installation assessment, which includes site inspection, record searches, identifies bases with closed, potentially hazardous waste sites.
• Existence of contaminants affecting environment are confirmed.
• Teach developed or advanced methods to solve problems if necessary.
• Remedial action designed and executed.

Findings from Sacramento Army Depot indicate continued heavy use of laborers (who earned 45 percent of gross pay) and operators (who earned 40 percent of gross pay). Other crafts at the site included electricians (1 percent of gross pay) and asbestos workers, drillers, and drivers who each earned less than 1 percent each of the gross pay. Sacramento Army Depot had a higher than average use of overtime as a percent of total hours (18 percent). The average hourly rate at Sacramento Army Depot was the second highest in the RRA data base at $23.85.

Data received from Concord Naval Weapons Station through a site visit and correspondence indicated that rates of pay for laborers and operators were about $30 per hour. Unfortunately, the RRA data base does not contain enough data from DOD sites to make any concrete generalizations. Interviews conducted with personnel involved in remediation activities at Concord Naval Weapons Station, McClellan Air Force Base, Hunter's Point Naval Shipyard, and Pillar Point Air Force Base, did not contradict earlier findings. One explanation of higher than average earnings could be that each of these sites is located in California. Data from these sites were used to strengthen projections on containment, USTs, and excavation remedies.

5. Underground Storage Tanks

Background on UST Sites

There are millions of underground storage tanks (USTs)³⁵ across the United States. USTs -- which store gasoline, crude oil, and hazardous substances -- are managed by RCRA Subtitle I. If they contain hazardous waste they are regulated under Subtitle C of RCRA. Over 1.6 million underground storage tanks are subject to federal regulation, with about 91 percent containing petroleum products and approximately two percent containing hazardous material.³⁶

Landfilling technologies comprise over half of the technologies selected to treat petroleum contaminated soils from USTs -- followed by in-situ treatment, thermal treatment, and bioremediation. About 40 percent of all UST cleanups involve innovative technologies -- most commonly soil vapor extraction, in-situ bioremediation, and thermal desorption. Technologies currently being used to manage petroleum contaminated soils at USTs are divided into three categories: ex-situ management, in-situ management, and groundwater management. In-situ technologies used are soil vapor extraction and bioremediation -- primarily for the management of groundwater, free product recovery, and pump and treat. Ex-situ technologies are used for low temperature, thermal strippers; hot and cold mix asphalt plants; for landfilling; for land treatment; and for stabilization and solidification.

Findings About UST Sites

Labor Mix

There are many descriptions of the work process for UST removal and remediation. The job usually requires the skills of laborers, equipment operators, and truck drivers. A composite of these skills may be combined into the work of an individual. The mix may change based on whether remediation requires removal or closure, the size and number of tanks, whether the tank is covered with earth or cement, etc.

According to a 1994 study,³⁷ the costs to a contractor of removing 18 USTs from a west coast military base (and closing five more in place) was $1,376,000 million, or, on average for removal only, $76,444 per tank. Of this total, 29 percent of the removal of tanks, piping, and off-site treatment and disposal was spent on labor.³⁸ The removal task itself was estimated to take 282.5 days³⁹ -- with 32 percent of the time for a site manager, 32 percent for an equipment operator, 32 percent for a laborer, and 4 percent for other workers. The off-site treatment and disposal was expected to utilize laborers and drivers, and perhaps operators as well.⁴⁰

Timing of Cleanup

Tens of thousands of underground storage tanks have been removed or closed, and cleanup of the sites completed. In FY92 alone, almost 29,000 UST cleanups were completed. From September 1994 through September 1995, the cleanup of over 40,000 USTs was completed. But, there appear to be significant differences across the country as to the ratio of completions to active tanks. Across the country there is an average of 12 percent completions from the universe of active tanks. It is as high as 16 percent in Region VIII and as low as 8 percent in Region VI and 5 percent in Indian lands. The rate of cleanup seems unrelated to the total number of tanks, as the most significant in-roads in UST cleanup seem to be in Regions VIII and IV, the ones with the least number and highest number of active tanks, respectively.

6. State and Private Sites

Background on State and Private Sites

Only a small percentage of the nation's hazardous waste sites will be placed on the NPL. Cleanup of Non-NPL sites will require federal, state, local, and private actions. Some are federal facilities; most are being cleaned up under the jurisdiction of states and territories. The U.S. General Accounting Office has estimated that there are between 130,000 and 425,000 state sites that might need to be evaluated for possible cleanup action.⁴¹ As a result of a report by EPA in conjunction with the Association of State and Territorial Solid Waste Management Officials (ASTSWMO) with further work done by Kensington Systems, Inc.,⁴² there were over 20,000 sites in a state/territory data base -- with thirty-nine states and two territories reporting.⁴³ It is difficult to bring all state sites into one data base for analytic purposes. According to EPA, the differences in terminology and systems for tracking accomplishments by the States and Territories makes the "the development of conclusions a challenge."⁴⁴

Additionally, because not many private site remediation efforts have been completed and because private sector cleanups are voluntary and there is no public source for keeping historical data, little detailed information exists on these sites. Compliance standards which strongly influence remediation costs vary from state to state, and may not always be enforced. Therefore, costs associated with these Non-NPL sites may be somewhat inconsistent across projects.⁴⁵

During the 1990's, EPA commissioned Kensington Systems, Inc.⁴⁶ to develop a data system on State and Territory Environmental Restoration Activities. This led, in August 1994, to the "State and Territory System Documentation" data base. In all, data were collected for 22,902 sites, whose remediation work cost over $1.2 billion. Information collected included duration of projects, the cleanup costs for state and PRP sites, and predominant remedies used.

Findings About State and Private Sites

Review of the EPA/ASTSWMO data suggests the major focus of states in remedy choice is off-site containment, with 61 percent of reported sites choosing it as a predominant remedy. (See Table S.) In some states; e.g., South Dakota, Rhode Island, and Alabama, off-site containment was the major remedy at 90 to 100 percent of all reported sites. (See Table T.) A few states focused equally on other remedies, with, for example, 100 percent of California sites (in the ASTSWM0 data base) choosing site security as major remedy and 76 percent of sites in Kansas using on-site treatment. The major crafts in demand were, therefore, most likely laborers, operators, and truck drivers and/or railroad personnel. Most actions took less than a year and cost between $300,000 and $400,000.⁴⁷ Among 4,000 predominant remedies across 31 sites in the data base, containment methods accounted for 76 percent of all predominant remedies, with 80 percent of the containments being off-site. (See Table S.)

Duration of Activities. A majority of state response actions reported in the EPA/ASTSWMO/ Kensington data base took less than a year to complete. At least half the response actions for 19 of 27 states reporting on duration were of less than a year's duration. (See Table U.) For five states, 80 percent or more of response actions lasted less than a year.

There were sites, however, where cleanup lasted three years or more -- nine percent of the total number of sites for which duration was reported (562 of 5,904 sites). Over half of these "long-term" sites (306) were in New Jersey. Another 136 were in Illinois, Massachusetts and South Dakota. Thus, just four states accounted for nearly 80 percent of all the sites where cleanup lasted for three years or more. Nine of the 27 states reporting on duration had no sites at all with remediation lasting three or more years.

Duration varied considerably from site to site and from state to state. While two states, Colorado and Texas, had no actions lasting more than one year, in Florida 25 percent of the sites reported activities lasting more than 5 years. The data base, reporting on 22,902 sites, contains only 3,527 removal completions and 2,428 remedial action completions through 1992. (See Table V.)

Predominant Remedies By Site. The predominant remedy at 76 percent of the State/Territory sites was containment -- 61 percent off-site and 15 percent on-site.⁴⁸ At federal sites, a broader array of remedies were used. At Federal sites, for both remedial actions and removals, 84 percent (2248 sites) used four predominant remedies: on-site containment, 24 percent; site security, 22 percent; off-site containment, 21 percent; off-site treatment, 17 percent.

Remedy data in the EPA/ASTSWMO/Kensington study were reported for 15,990 sites in 31 states. Predominant remedies were divided into seven major categories:

• On-Site Treatment
• On-Site Containment
• Off-Site Treatment
• Off-Site Containment
• Population Protection
• Site Security
• Innovative Technology

Sixty percent of on-site containment involved surface capping, rarely with a slurry wall. The next two most frequently used on-site containment techniques were surface drainage control and soil cover. Sixty percent of off-site containment involved excavation and final removal to an off-site landfill.

At seventeen percent of all state sites, treatment, either on-site or off-site, was a predominant remedy. Five remedies accounted for 80 percent of on-site treatments: pump and treat with off-site discharge, pump and treat with on-site disposal, air stripping technologies, soil aeration technologies, and biodegradation (in that order of usage). Incineration, pH neutralization, component separation, leachate treatment, and thermal treatment were less frequently used on-site treatment remedies. (See Table W .)

OTHER ISSUES

CHOICE OF REMEDY

Labor Market Demand By Type of Remedy

Clearly, the labor mix at a hazardous waste site is significantly affected by the remedy chosen. The first major choice is between containment and treatment.⁴⁹ Containment usually focuses on the skills of laborers, operators, and truck drivers; while treatment remedies are more likely to require a broader involvement of other crafts as well. The operations and maintenance phase of treatment cleanup is likely to require a range of industrial workers and machine operators.

Treatment rather than containment is the mandated remedy of choice at Superfund sites, according to the 1986 SARA Amendments. Following the passage of that legislation, the percent of source control RODS with a treatment remedy, for Superfund remedial actions, rose from 51 percent in FY87 to 84 percent in FY90. But the trend has reversed since FY90, with the percent of RODS with a treatment remedy falling each year -- and the FY94 percent falling to 57 percent.⁵⁰ The reason for this fall? Many blame the uncertainty of Federal regulations for the unwillingness of the private sector to take a risk with innovative efforts.

Still, many innovative treatment technologies are being developed. For soil treatment, these include: bioremediation, dechlorination, vitrification, soil vapor extraction, and thermal desorption. For groundwater technologies, these include, air sparging, bioremediation, and surfactant flushing.⁵¹

At state remediation and removal sites the major remedy chosen has been containment, mostly off-site. To the degree that decisions and responsibility move toward the state, pre-SARA labor mixes and state site remedies are more likely -- with a probable return in national focus to containment rather than treatment.

Remedy choices across the DOE complex vary considerably. Because there is significant groundwater contamination at many sites, there is a necessary focus at least at these sites, on treatment; e.g., at Hanford, Savannah River, Rocky Flats, and Oak Ridge. Operations and maintenance for pump and treat associated with groundwater extraction systems may require time frames as long as 100 years.⁵² But at some DOE sites, the focus may be more on excavation and disposal, as at the Nevada Test Site or Idaho National Engineering Laboratory. Likely, there are combinations of remedy choices; e.g., at Fernald or Rocky Flats, where there is treatment and then off-site containment.

Because of the expense of transportation, excavation and disposal strategies are relatively costly. Incineration tends to be the most costly technology, but its performance is proven. Many in-situ technologies are relatively less expensive, but their applicability and effectiveness is less wide-spread. Unless a change in compliance standards or enforcement occurs, there will probably continue to be widespread use of proven, more expensive treatment options.⁵³

The sites studied serve as an illustration of the diversity of crafts required for a task. For example, across the sites, carpenters, cement masons, electricians, iron workers, laborers, operators, painters, and plumbers and pipe fitters were all involved in the construction of decontamination pads. Well installation required the labor of carpenters, laborers, operators, and drillers.

Labor Mix With Containment Remedies

Labor mix varies with different containment remedies. Methods of containment include multilayered RCRA caps, soil caps, slurry walls, surface water diversion through grading and dike construction, solidification and stabilization and more. Also a factor in the labor mix is whether the final containment is on-site or off-site.

Some of these methods will require more diversity in the labor mix than others. For example surface water diversion through grading and dike construction is likely to be more diverse in the mix of crafts -- using pipe fitters, plumbers, perhaps cement masons, as well as laborers and operators, whereas excavation and off-site landfilling is likely to only require, laborers, operators, and drivers.

Landfills

The RRA data base contains certified payrolls and daily labor logs from four landfills -- Lipari and Lone Pine in New Jersey, Moyer Landfill in Pennsylvania and New Lyme Landfill in Ohio. All four landfill sites are on the NPL list. Lipari, Moyer, and New Lyme are Fund-lead sites, Lone Pine is a PRP-lead site.

Landfill Remedies. The presumptive remedy for landfills is containment because treatment of landfills is often impractical due to the size and/or the combination of contaminants within. Containment may be accomplished by capping, leachate collection and treatment, landfill gas treatment, and/or institutional controls.

Out of forty-four possible remedies for thirty landfills studied by EPA,⁵⁴ the multi-layer cap was considered, and passed through the feasibility study screening most often (eighty-nine percent). It was chosen as the remedy sixty-four percent of the time (more than any other technology). Some other alternatives considered were: clay caps, asphalt caps, concrete caps, synthetic caps, soil covers, chemical seals, and slurry walls. Clay capping passed screening fifty percent of the time, and was the selected remedy twenty-five percent of the time that it was considered. Twenty-eight of the technologies, including asphalt and concrete caps, never passed through the screening during the feasibility study.

The amount of money and number of hours involved in each of the four sites studied by RRA differed; nonetheless, the tasks at each were similar: line and cap the landfill so as to control contamination by hazardous materials. The technologies differed to some extent -- at Lone Pine a CLAYMAX cap was used, as was a carbon treatment system to treat groundwater/leachate both chemically and biologically. At Lipari a multilayer cap was used, as well as a batch flushing treatment system. Moyer used soil clay capping, and New Lyme used a RCRA cap, and used biological disc, sodium hydroxide precipitation, and granular activated carbon to treat the contaminated groundwater and eliminate leachate. However, overall remediation strategies -- capping the landfill and implementing groundwater/leachate collection systems -- did not differ substantially.

Landfill Jobs. There may be many categories of labor involved at each landfill site -- often more than a dozen. At Lipari a mixture of skilled crafts accounted for a significant amount of hours, while at Lone Pine laborers accounted for 34 percent of the person-days, with very few other crafts present. At Moyer, laborers, operators, and drivers accounted for 94 percent of the hours worked. At New Lyme, although many crafts were present, laborers, operators, and drivers together accounted for 70 percent of the hours worked.

The proportionate share of each of the major job categories differed from site to site. Laborers accounted for 18 percent of total hours at Lipari, 34 percent of person-days at Lone Pine, 50 percent of total hours at Moyer, and 21 percent of total hours at New Lyme. Operators accounted for 9 percent of work at Lipari, 23 percent at Lone Pine, 38 percent of total hours at Moyer, and 30 percent of total hours at New Lyme. The proportion of drivers to laborers to operators was approximately the same for three of the sites, but was much higher at New Lyme, where drivers had a much higher presence. The difference in overall percentages may be the result of the large number of other crafts working on a variety of tasks at Lipari, thus bringing down the overall percentage of labor attributed to drivers, laborers, and operators.

Why the differences? It is not clear the extent to which the job categories are defined differently at each site. The major difference probably has to do with the differences in the design of the plants. At Lipari the plant also operated as office space for O&M workers, while at Lone Pine the plant was designed as more of a raw structure. Thus, more finishing tasks were needed for the plant at Lipari, and more specialized craft workers were present. In addition, at Lipari a major sewer line was built, thus requiring a high percentage of plumbers and pipe fitters.

Landfill Pay. Davis-Bacon wage rates prevailed at all the landfill sites studied. Average hourly wages at Lipari ranged, among the predominant job categories, from $16.60 for laborers to $23.52 per hour for mechanics. At Moyer average hourly rates ranged from $17.57 for laborers to $21.20 for operators. Average hourly wages at New Lyme ranged from $11.00 for drillers to $20.53 for electricians. ( See Table F.)

Excavation and Hauling

Excavation and/or hauling were the major activities at four California sites studied. Whereas one might have expected somewhat parallel staffing of these activities, such was not the case. Though all four sites -- a former battery plant, Embarcadero in San Francisco, Hunter's Point Naval Shipyard, and Pillar Point Air Force Station -- utilized the services of truck drivers, laborers, and operators; actual labor demand was affected by the following issues:

• If the hazardous material needed to be taken out of state, then rail was often involved and a number of additional laborers were needed to line rail cars. (For example, at Embarcadero and Hunter's Point.)
• If material was stockpiled at an intermediate point, awaiting its final destination, then additional operators were needed. (For example, at Embarcadero and Hunter's Point.)
• The number of truck drivers is highly dependent on the length of the haul and complexity of the truck.
• In some instances a scale operator may be needed. (For example, at Hunter's Point.)

EMERGENCY RESPONSE

Worker protection standards governing hazardous waste work include requirements pertaining to emergency response activities. However, the hazardous waste work employer has options with regard to how to respond to these emergency response requirements. A hazardous waste site cleanup contractor can employ an "emergency action plan" as the site emergency response program. In this instance, an emergency response system is applied to the site which involves recognition, issuance of an alarm, evacuation, and notification of an off-site emergency response organization; HAZMAT, fire and rescue, and/or emergency medical. On the other hand, that employer may organize, train, and utilize an on-site emergency response crew to deal with site emergencies. This is often referred to as a collateral duty emergency response team or, in the industrial setting, as a fire brigade.

"Emergency Action Plans" are the current choice for the majority of hazardous waste activities associated with removal, remediation, O&M, USTs, and like activities. In these cases, employers, in compliance with OSHA/EPA standards, coordinate emergency response activities with off-site response organizations; the local community emergency response organization is called in when an emergency situation develops. In this case, site workers may still need to be trained as first responders who may recognize the incident and need to report it.

RCRA/TSD sites far more frequently utilize collateral duty emergency response teams. In these cases, currently employed workers will be required to take additional hazardous waste training, in accordance with the OSHA/EPA regulations governing emergency response and the training of such responders. The actual number of workers engaged in emergency response in these instances will parallel the increased numbers of workers engaged in RCRA-TSD corrective actions. That is, where collateral duty emergency responders are utilized, there will be an increase in emergency response training demand. For those RCRA-TSD corrective action sites who develop "emergency action plans" the local community emergency response organization would be utilized as discussed previously.⁵⁵

The demand for emergency response clearly increases as the volume of hazardous waste work increases. Additional on-site staff may be needed at hazardous waste locations. Additional burdens are also placed on existing emergency response personnel -- most often fire, police, and emergency medical employees of the public sector. The supply of emergency responders in the public sector is relatively inelastic, so an increase in HAZMAT incidents is likely to cause a parallel increase in the use of overtime.⁵⁶

Findings on Emergency Response from Data Studied

Analysis of data from five data bases in three states (Arizona, California, and New Jersey)⁵⁷ documented the thousands of reported hazardous materials incidents that occur each year across the United States. Each reported data somewhat differently. While the level of under reporting and early reporting may significantly understate the number and severity of incidents, at least seven important facts do emerge:

• Transportation corridors are the location of a significant number of hazardous materials incidents. In Arizona, 48 percent of incidents were along transportation corridors;⁵⁸ in California, 33-37 percent of incidents were on ground transportation routes and involved "transportation" property.
• Emergency responders as well as facility workers and community residents are at significant risk of injury and fatality in the event of a hazardous release. In California in 1990 and 1991, emergency responders accounted for 12-15 percent of those injured. The AHMIR reports showed that of those exposed to an incident, responders were 32 percent of those contaminated and 5 percent of those injured.
• HAZMAT emergencies are often not contained quickly. In New Jersey, for example, of 2000 reported incidents over a nine-month period, only 23 percent had been contained at the time of reporting.
• There were injuries at five percent of the reported incidents in New Jersey. Fatalities were reported for one tenth of 1 percent of incidents. In California, both in 1990 and 1991, there was an injury/incident ratio of just over 0.2 (700 injuries in 3300 incidents in 1991 for example). Fatality/incident ratios in 1990 and 1991 were 0.001. In Arizona the ratio of injuries to incidents was 0.15-0.33 and the ratio of fatalities to incidents was 0.005 to 0.014, depending on the data base studied. The ratio of exposures to incidents reported in fire department records in Arizona was 0.76.
• Public sector emergency responders were not always present at incidents. While in New Jersey, fire services responded to 77 percent of the incidents and police also responded to 77 percent of the incidents; for 18 percent of the incidents neither fire nor police were on the scene when the incident was reported to DEPE (the state environmental protection agency).
• For 4 percent of the New Jersey incidents, evacuations of the public were required.
• Improvement in response time is needed. The U.S. Environmental Protection Agency considers response times less than 15 minutes as "timely." Statistics from the Arizona Fire Department for response time in 1993 showed average response time of 19 minutes though the median response time was 10 minutes. For forty-eight percent of the incidents the response time was more than 15 minutes -- with response times ranging from 0 to 91 minutes in all but one instance, when the response time was 254 minutes.

Hazardous materials incidents in 1993 and 1994, as studied by ATSDR in nine states,⁵⁹ were responsible for 11 fatalities, 1446 injuries, 457 mass evacuations, and 3125 episodes of accidental release. These HAZMAT incidents occurred in commercial/industrial areas over 60 percent of the time. One-fourth of the incidents occurred during transportation. Most frequently released were industrial chemical gases, herbicides, and acids. Two-thirds of the time injuries were to workers, with nearly 20 percent affecting residents and 14 percent affecting emergency crews.

Additional Issues

Where clean up and related activities occur at DOE sites, there may be special additional emergency response requirements associated with those wastes present at nuclear facilities. In these instances, additional emergency response planning, preparedness and response actions will be governed by NRC, FEMA, DOE, and perhaps others. If these special nuclear waste activities increase, one might expect an increase in the need for emergency response personnel and training, although the dimensions of such an increase may be small and difficult to qualify.

There are some nine Federal Agency standards requirements for emergency response activities. Recently, the "ONE PLAN," the National Response Team's Integrated Contingency Plan Guidance was published in the Federal Register (61FR 28641). This Plan provides guidelines which, if followed, would achieve compliance with all current Federal emergency response requirements.

The hazardous waste worker protection standards by OSHA and EPA are the only agencies whose standards apply to less-than-full-time emergency responders; i.e., volunteers. To the extent that hazardous waste clean-up activities are conducted in places where the emergency response needs are met by volunteer emergency response organizations, special attention to the training and equipment needs of those volunteer departments will continue to be needed.⁶⁰

Demand for Labor: Emergency Response

An analysis done for the Occupational Safety and Health Administration, estimated that there were 28,000 local fire departments, 22,000 industrial in-plant emergency response teams, 750 commercial (private) hazardous material response teams, and 200 public hazardous material response teams.⁶¹

The average number of workers assumed to be involved in response was 10 for local fire departments, commercial response teams and public response teams and eight for in-plant response teams.⁶² Data from two surveys reviewed by an OSHA contractor found that the ratio of fire brigade members to total employees in manufacturing was approximately 45:1.

Staffing needed for a HAZMAT team includes at least 2 entry staff, 2 backups, and 1 safety officer. In addition, emergency medical people must be on standby within a certain amount of reaction time to the incident.

When using self contained breathing apparatus, there must be a minimum of two people inside and two outside. The two inside must be trained to technician level, the two outside must be trained to operations level. Also, one of those on the outside will serve in a dual role; also functioning as incident commander. The incident commander decides who else to call in. Most often it is a snowball effect, with one agency calling another, and so on, until there is sufficient response.

Emergency Response at Nuclear Facilities Has Additional Requirements and Notification Procedures

There are four levels of Emergency Response at facilities for which the Nuclear Regulatory Commission has oversight:

1.  Unusual Event: No release of radioactive material
2.  Alert: Minor incident
3.  Site Area Emergency: Release that is within the boundary of the facility
4.  General Emergency: Release that is beyond the boundary of the plant

The emergency planning zone is a 10 mile radius of the facility. At Perry Nuclear Power Plant in Ohio, this includes three counties, with one taking on the role of Emergency Operations Center.⁶³ They train, out of three counties, 3500 - 4000 people a year. This involves, one to four hour emergency responder training, with more time if necessary. Those trained include school bus drivers. (who get two to three hours of training).

Every two years there is a disaster drill, evaluated by the Nuclear Regulatory Commission and the Federal Emergency Management Agency.

ENVIRONMENTAL JUSTICE AND ECONOMIC OPPORTUNITY

Many hazardous waste sites are located in close proximity to low-income communities made up of African-Americans, Hispanics, Native Americans, and other people of color. While there is not full agreement on the reason for this inverse relationship between race, economic status and environmental quality, there is general agreement about the need to accelerate the process of hazardous waste clean up, especially in underprivileged communities from which departure is constrained by economic reality.⁶⁴

The data derived from this study help to identify where hazardous waste workers live. Some believe that one important part of the environmental justice equation is that those whose communities have been most negatively affected should have the largest possible opportunity to gain employment in the cleanup.

At an individual community level, the infusion of thousands, even millions, of dollars allows for the growth of local economic base when those dollars stay within a community -- creating potentially important sources of funds for both the HAZMAT worker and for those from whom these workers purchase.

In many instances local individuals form a large segment of a cleanup work force. At the Shiprock UMTRA site, located on the Navajo Nation, for example, over 98 percent of the gross pay from certified payrolls was earned by workers who lived within the Navajo Nation. (See Table 2 for Shiprock.) Nearly 50 percent of the total was earned by those who lived within the town of Shiprock itself. Thus, individuals in a town of less than 7700 people earned over half a million dollars in the cleanup of their community, and many learned new skills as well.

A more detailed review of the communities in which workers reside was done for Lipari Landfill. In the communities where Lipari workers resided, Lipari workers were among the higher wage earners. Though census data is in medians and certified payroll data is in averages, the trends are still clear. As the percent of non-white residents in communities surrounding Lipari goes up, median hourly income generally goes down. (See Chart 2.) Regardless of racial composition, average hourly wages at Lipari were consistently higher than earnings throughout the community. The wages earned by Lipari workers did not alter significantly with racial make-up of communities where workers resided.

Workers at Lipari provided above average earnings across many New Jersey communities. While many New Jersey residents living in relative proximity to Lipari gained employment, one town just adjacent to the site, Glassboro, with an unemployment rate over 10 percent, received only $15,000 in pay from a studied payroll of over $4 million. Conscious efforts to include community residents in the work force at Lipari might have altered that.

There seems, however, to be no pattern in the level of skills and racial composition of workers within a community; i.e., workers from communities that were more largely non-white were as likely to provide iron workers and operating engineers as general laborers and truck drivers.⁶⁵

The demand for workers to clean up hazardous waste is very large -- through the year 2010, well over five billions hours of work will be required. The jobs are in every state and span job skills of the construction, industrial, transportation, and emergency response work forces. Any predictions of the future require that assumptions be made about the future. Assumptions made in this section of the study are discussed below, as are outside sources and the methodology used. Projections are based on real experience at hazardous waste sites over the past decade or more. The detailed experience-based approach of this study provides perspective, depth and credibility to the projections made. Using the same data, other researchers might use different assumptions that could result in other arrays of projections.

A. PROJECTIONS OF THE OVERALL NUMBER OF WORKERS NEEDED TO COMPLETE MANDATED HAZMAT CLEANUP

The RRA data base provides experience-based information on average hourly earnings and craft mixes at remediation sites. Using these data, along with data from an extensive literature review on the cost of remediation activities, projections were made on the number of jobs generated by hazardous waste cleanup work. The process described in this chapter yields an overall demand of 3.4 million job years, with approximately 5.4 billion hours worked on-site by remediation workers.

In addition to these direct on-site jobs, many others will be employed to complete a wide variety of tasks:

• Other craft labor jobs are generated in phases other than remedial action; e.g., well drilling during site investigation;
• Other craft jobs exist for ex-situ treatment and containment or for transportation of hazardous waste off-site;
• Many more direct jobs are generated for engineers, scientists, laboratory technicians, field technicians, and government officials throughout the cleanup process;
• Jobs are also generated for a wide array of other tasks -- from suppliers of utilities, to manufacturers of equipment and parts, as well as through a general multiplier effect in the economy.

Assumptions. There are many assumptions that affect projections. The RRA data base describes labor mix and wages from an array of completed hazardous waste site tasks. To make projections from the RRA data base, a number of assumptions had to be made.

• Number of Sites. New hazardous waste sites may be found. Criteria for listing sites may change and some sites, once on a cleanup list, may be dropped. As this number changes, the number of jobs needed to complete cleanup will change.
• Amount of Contamination. Sometimes the true extent of contamination is unknown until work has begun. When earth moving begins, there may be new discoveries that change estimates of the types and/or volume of contaminants. Researchers in one study⁶⁶ found that the volume of contaminated soil at 56 percent of the sites studied had been underestimated. This led to a 43.5 percent overall increase in the volume of soil which needed remediation.
• Choice of Remedy and Technology. The choice of remedy and specific technology used at a site clearly influences the cost of cleanup and potentially influences the labor mix as well. Containment is likely to cost much less than using an innovative technology to treat contaminated soil.
• Duration of Cleanup. Although one would expect the length of cleanup action to influence the cost with longer cleanups costing more money, this is not always the case. For example, when examining state removal sites no correlation was found between length of cleanup and cost. While one removal site took 57 months and cost $36,850 another removal took 7 months and cost $214,000. (See Table X.)
• Duration of Operations and Maintenance. The longer operations and maintenance activities continue the more expensive the site cleanup will be. Because O&M activities are just starting at some sites and have not yet begun at others, actual duration of O&M activities, as well as total costs are still uncertain.
• Choice of Equipment. The equipment used at hazardous waste sites clearly influences the labor intensity of work. At sites where hand tools are used significantly there are smaller units of earth moving at a time than when backhoes or bulldozers are used or when forklifts and cranes are added to a site. When contamination is close to a building, a more labor intensive effort may be necessary than when the contamination is in an open field because small tools may be needed to get close to foundations.

Use of the Word "Job". All calculations for jobs in this text are for a full-time job lasting one year. For construction labor, a full-time job-year is assumed to be 1500 hours of work. For operations and maintenance labor, a full-time job-year is assumed to be 2000 hours of work.

Discounting of Costs. Many studies which project the cost of hazardous waste cleanup discount those costs. Authors struggle to determine an appropriate discount rate to use, so that costs may be expressed in an accurate present value. In this study, only undiscounted costs are sought and used. This is because the research interest is not in present value, but rather in the actual dollars to be expended -- dollars which in turn reflect on the number of jobs being supported from those expenditures.

Sensitivity. A formal sensitivity analysis is not part of this report, but sensitivity analysis was performed throughout the research portion of the work. As additional sites were added incrementally to the data base, summary conclusions were not particularly sensitive. When assumptions were made for the purpose of making projections, the sensitivity associated with those assumptions was great, and this sensitivity is referred to throughout the text. Assumptions about cost of cleanup, the percent of cleanup costs accruing to remedial action and to construction labor, and the number of hours assumed to represent a full year of work, all have significant impact on the overall projections.

Uncertainty Associated With Making Projections. Making forecasts always involves uncertainty. Because this data base makes its projections from a diverse array of sites across the nation, based on actual work experience, many of the problems associated with extrapolating from one site -- and perhaps only the plans for cleaning up that site -- are eliminated. Nonetheless, the following issues are still obstacles to the challenge of making projections:

1 The degree of cleanup is dependent on the political process and policy choices, which affect budget, timing, research into innovative technologies, designation of responsible parties, intensity of cleanup, etc. The cleanup of individual sites is under the jurisdiction of varying regulatory bodies, each with different mandates and cleanup requirements.

2. There are fundamental philosophical/political issues to be resolved. There is a major question of whether cleanup should be to resolve imminent danger (at a relatively low cost) or whether future generations should be protected, and hence more expensive groundwater or soil treatment should be used. To the degree that decisions and responsibility move toward the state, pre-SARA labor mixes and state site remedies are more likely; i.e., more containment and less treatment, and most of the containment will be off-site.

3. The diversity among state programs is great. Different states have different patterns of choosing remedies. Lying behind policy may be issues as diverse as unfunded mandate concerns, avoiding future liabilities, restoring land to productive use, avoiding the possibility of exposing workers or surrounding communities to risk.

4. To the extent that monies are expended on the front end to prevent pollution and find alternative ways to limit hazardous waste and dispose of it safely, waste management costs as well as the costs of cleanup will be diminished.

5. As the environmental remediation industry pursues R&D associated with more effective and more cost-efficient and innovative remediation technologies, the types of work tasks required may change.

6. Data are incomplete and inconsistent. Estimates have largely been for federal costs alone, with nonfederal costs excluded.⁶⁷ When the U.S. General Accounting Office surveyed federal agencies to see how far along they were in estimating costs, they found that officials at most agencies were uncertain about critical aspects of cleanup, such as the nature and extent of contamination problems and type of cleanup strategies needed.⁶⁸ Hence, GAO decided that it was impossible to estimate either costs or labor market needs.

When necessary, this study relies on the findings of past research efforts. These findings are not always consistent and the process of making projections required choices. For instance, many of the projections of the total cost for environmental restoration and waste management at DOE sites are estimated at $230-240 billion, but the estimated cost of cleanup at Hanford alone is $240 billion or more.

Another problem with projections is making a sensible estimate of the number of sites requiring cleanup. EPA in 1994 estimated the cost per site of NPL site cleanup to be $30.7 million with a total cost of $42 billion. This suggests the cleanup of 1368 sites, but EPA's Assistant Administrator Elliott Laws, also in 1994, suggested in a response to Congress over 1700 sites and the EPA Inspector General in 1994 estimated 3000 sites. In 1994 the GAO was publishing the Congressional Budget Office's best estimate of 3300 NPL sites.

Projections for HazMat Jobs 1990-2010

The base for projections is the actual experience at hazardous waste sites, as documented through certified payrolls. From this data-based picture of several past and present job sites and the labor used, projections for future labor demands are made. But, in order to make those assumptions, other data and a series of assumptions are necessary.

An estimate of the number of sites that will require cleanup and a cost per site for the array of site types and locations -- based on type of contaminant, amount of contaminant, and technologies to be used is needed. This report, because the focus was the collection and analysis of certified payroll data, relies heavily on the cost estimates available through the responsible government agencies or made by consultants and contractors.

One needs to assume the number of labor hours that comprise a full-time work year -- a number that is usually significantly lower in construction than for industrial labor. One must also estimate the share of remediation dollars that go to construction payroll and the share that go to operations and maintenance. And, finally, one must estimate the time frame in which the cleanup dollars will be spent, by category of cleanup. Once these assumptions have been established, projections can be calculated.

For the purposes of this study, the following assumptions were made:

Estimates of the Number of Hours that Comprise a Full Time Equivalent (FTE) Job for a Construction Worker for a Year. In some parts of the country, a construction worker has little expectation of work for more than 1000 hours a year. A forty hour, 52 week a year job provides 2080 hours a year of work. Many in construction consider 1400, 1500, or 1600-1800 a reasonable benchmark for a full year of construction work. The Current Population Survey, for 1992 and 1993 found the average construction worker employed 1981 and 1994 hours per year, respectively. Projections in this report assume a 1500 hour per year work year -- a floor that gained acceptance among most building trades experts interviewed.

Estimates of the Share of Remediation Dollars That Go to Construction Payroll. Estimates of the share of remediation dollars that go to construction payroll at NPL and other hazardous waste sites is estimated to be 15 percent; based on the following assumptions and expert opinions:

Projections for Nationwide Hazardous Waste Cleanup

The labor market projections in this study are based on the findings reported in Section II, along the cost estimates and assumptions found at the beginning of Chapter III. The individual category projections are explained in the pages that follow. For summary and overview purposes the full projections are presented first. (See Table 5 for "best" estimates and Table 6 for a reasonable upper bound.)

The 3.4 million expected job years of labor for cleanup yield an estimated 5.4 billion hours of work -- based on an estimated 2.9 million construction job years (1500 hours per job year) and 500,000 O&M job years (2000 hours per job year).

Table 5

Using different cost estimates, from a study done for DOE, the following projections were made:

Table 6

Estimated Cost of Cleanup

There are many estimates on the cost of cleaning up the nation's hazardous waste sites. All are projections based in significant uncertainties. Thousands of sites have yet to be thoroughly characterized, and once cleanup actually begins, there may be many new discoveries of additional types or amounts of contaminants. Federal, state, and private budgets, as well as public policies, may change.

Despite wide ranges in cost estimates, analysts agree that the remediation of hazardous waste sites will cost many billions of dollars. According to a widely cited 1991 University of Tennessee study,⁶⁹ the total cost of cleaning up U.S. hazardous waste sites -- NPL, RCRA, DOE, DOD, UST, and others -- will be $752 billion. (See Table 7) According to a comprehensive DOE-sponsored study,⁷⁰ the total cost will be $983 billion. (See Table 8) These two studies are broadly viewed as the most comprehensive studies of costs across hazardous waste activities. Each makes estimates for the cost of cleaning up sites by sector of responsibility; i.e., NPL, RCRA, DOE, DOD, Underground Storage Tanks, and State/Private sites. No other comprehensive study reviewed by RRA included such detailed cost estimates.⁷¹

Table 7

The estimated cost of cleanup for Superfund sites varies greatly -- with ranges in this study from $37 billion (Resources for the Future) to $352 billion (University of Tennessee). DOE estimates its cleanup costs at approximately $230 billion, even though estimates are that cleanup at the Hanford site alone will cost at least that much and maybe even as much as $500 billion, over 75 years.⁷² Estimates on remediating DOD sites range between $25 billion (DOE) and $30 billion (University of Tennessee). UST site cleanup estimates range from $49 billion (DOE) to $67 billion (University of Tennessee). Other cleanups -- including state, private, and non-NPL -- are estimated to cost between $30 billion (UT) and $153 billion (DOE).

The UT and DOE studies each developed numbers in mutually exclusive categories. But, it is not always clear into which category a site falls. A site might, for example, be under the jurisdiction of DOE, be listed on the NPL and also be regulated by RCRA. In this case, DOE would be responsible for all costs incurred in the remediation; including those which are regulated under CERCLA and RCRA.⁷³ The categorization of sites also affects the numbers for each type of activity. Sometimes, for example, the line between environmental restoration and waste management at DOE is unclear; e.g., the classification of underground storage tanks at DOE's Paducah site are environmental restoration, but at the DOE Hanford site, the tank farm activities are classified as waste management.

Estimated Cost of Operations and Maintenance

Major types of O&M activities include pumping and treating groundwater and in-situ treatment of soils and groundwater. For groundwater remediation, operations and maintenance costs are of primary concern because pump and treat systems operate over long periods of time. One study⁷⁴ suggests that while costs for groundwater remedial action are generally based on a 30-year period for O&M expenses, that remedial time frames for groundwater are underestimated by at least a factor of three. Costs include such post-installation expenditures necessary to provide for the continued effectiveness of remedial action as labor and material costs, sampling and analytic fees, and administrative costs⁷⁵.

Because few cleanups have reached the O&M stage, many O&M costs are still uncertain. One major study estimated that for each site category -- NPL, DOE, DOD, RCRA, Non-NPL, and Underground Storage Tanks -- O&M cost equals approximately 16 percent of total remediation costs. For NPL sites, EPA has estimated O&M to be 18 percent of total costs incurred during cleanup of an NPL site.⁷⁶ For the purpose of this study O&M costs were more conservatively estimated to equal 16 percent of the total remediation cost.⁷⁷

B. PROJECTIONS OF LABOR MARKET NEEDS BY CATEGORY OF CLEANUP SITE

For each of the six major categories of cleanup sites discussed above, separate projections were made -- based on data specific to that category. These projections are presented below for Superfund, RCRA, DOE, DOD, UST, and State/Private sites.

1. Total Estimated Job Generation From NPL Site Cleanup

Estimating job requirements for NPL site cleanup is based on a number of important data sources and research assumptions. (See first part of Section III.) Beyond data finds from certified payrolls, it is necessary to estimate the number of sites and the cost of cleanup. Assumptions about the number of hours that represent full time equivalent job, the percent of total cost represented by construction labor on O&M.

Estimates on the Number of NPL Sites

There are many estimates of the total number of hazardous waste sites that will obtain Superfund status and be remediated. The three estimates listed below provided a range based on important and responsible sources:

All estimates need to be treated as just that -- estimates. The cost estimates used here include different arrays of numbers. Efforts have been made to use undiscounted numbers because it is actual dollars expended that aid in projected wages and jobs. Some are in terms of dollars of different years than others. The University of Tennessee and Department of Energy estimates are for 30 years (through 2020). The CBO estimates assume that cleanup lasts for more than 50 years. DOE is now using a 75 year time frame at its sites, just for remedial action completions.

Cost of Cleaning Up NPL sites

There are many estimates for the cost of Superfund cleanup. The ranges of costs reviewed in this study are from $37 billion to $352 billion. (See Table 9.) EPA's best estimate for the average cost per site is $30.7 million. (See Table 10.)

Averages, however, may be misleading. At non-Federal NPL sites, for example, a small number of sites consume the lion's share of total NPL capital expenditures -- with 16 percent of the operable units accounting for over 60 percent of all cleanup costs. The majority of projects (69 percent) have capital costs of less than $10 million and 38 percent have capital costs of less than $3 million.⁸¹ Full cost estimates per site, made in 1995 by Resources for the Future (RFF)⁸² ranged from $10.7 million for TNT processing sites to $1704 million for mining sites. The average cleanup cost for landfill sites, the most common type of NPL site, was estimated to be $23 million; leaking container sites $34 million, and radiological tailing sites $75 million. The RFF average cost for NPL cleanups is the lowest of those surveyed in this study, $29.1 million. (See Table 9.)

O&M costs vary by type of remedial technology, with ongoing pump and treat of groundwater being a common O&M activity. O&M functions, when required, may continue for several years and represent activities that are distinct from the cleanup effort. O&M costs depend on many factors including the type of contamination at the site, the different media that must be treated, and the remedial technology selected. One DOE study asserts that O&M costs are positively related to the size of the site when containment strategies are employed, but probably not related to site size when a treatment strategy is utilized.⁸³

Table 9

This study, for cost estimating purposes, focuses on the University of Tennessee is best "estimate" of $151 billion -- a figure cited often in the literature and a low middle number among the cost estimates reviewed. In 1991, the University of Tennessee's Waste Management Research and Education Institute issued a series of six studies on hazardous waste cleanup addressing cleanup costs for the wide array of NPL sites, RCRA sites, DOE sites, DOD sites, underground storage tank sites, and state and private sector sites. The University of Tennessee study was chosen because only it and the DOE-sponsored study were comprehensive -- and of the two, the University of Tennessee was more conservative in its costs estimates. The University of Tennessee estimates are frequently referenced and a baseline for the continuing research by many groups, ranging from Resources for the Future to GAO.

EPA estimates that 73 percent of the costs at a given NPL site are for remedial action work -- work which most intensively involves construction labor. For the portion of total NPL cleanup costs that, on average, is allocated to each major function, see Table 10.

NPL Cost Estimates from the Literature. The literature provides many estimates for costs associated with cleanup of Superfund and other hazardous waste sites. Results from several of these studies have been used to create a range for determining a responsible, conservative cost estimate to use as the base for projecting the numbers of jobs generated at Superfund sites. Cost estimates from five major studies are the focus of estimates in this report:⁸⁴

When using University of Tennessee's best estimate, over half a million job years are generated as a result of construction labor activities in remedial action. After reviewing sixteen different estimates of the cost of Superfund cleanup, and a number of other assumptions, projections for job generation were made (See Table 11.) ranging from a low of 130,000 to a higher of over 1.2 million.

Table 11

The first seven estimates depend on projections for total cost of NPL cleanup -- three from a 1991 study by the University of Tennessee, three from a 1993 study for the Department of Energy, and one by the Chemical Manufacturers Association (CMA). Based on these seven estimates the number of full time construction labor jobs generated as a result of Superfund cleanup ranges from a low of 129,000 to a high of 1.2 million. A middle and still quite conservative range might be 500,000 to 600,000 building trades job years over a 30 year period.

The remaining nine estimates are based on two sources estimating the cleanup cost per site, each with further estimates based on the number of NPL sites to cleanup. Two cost-per-site estimates are a University Tennessee estimate of $50.3 million per site and a Brookings Institution and Resources for the Future estimate of $29.1 million per site.⁹² The latest average cost per site comes from EPA's April 1995 release of the NPL which estimates the average cost per site at $30.74 million. The three estimates of total number of NPL sites to be remediated are: a very conservative estimate of 1285 based on the number of actual NPL sites in April 1995,⁹³ a 3300 estimate of the CBO,⁹⁴ and an EPA estimate of 4300 based on an addition of 3000 of the 6467 sites in the agency's evaluation backlog in 1994.⁹⁵ The CBO estimate of 3300 had a range of 2300 to 7800 sites. The high range was excluded so as to keep estimates conservative.

Using actual average hourly earnings for each craft from the RRA data base, rather than the average hourly rate for all Superfund sites in the data base, causes the variance of total number of remedial action jobs found in Tables 11 and 12.

Table 12

Projections of NPL Operations and Maintenance Jobs

An estimate for O&M costs as a percentage of total cost is from a 1993 study supported by DOE,⁹⁶ which estimated O&M costs across all sites, on average, at 16 percent of total cost. This 16 percent estimate is used for projections in this study.

Operations and maintenance costs associated with a Superfund site can be significant. Annual O&M costs were available for six of the eleven NPL sites studied in this paper. Costs ranged from only two percent of capital costs annually at New Lyme Landfill to 56 percent at Hollingsworth. (See Table 13.) Based on an assumption of 20 years of O&M per site (many estimates are based on 30 year or even 75 years), the total cost of operations and maintenance at two of the six sites is greater than total capital costs.

Table 13

¹ Of the 1 remedial action that requires O&M
² Based on annual O&M of $252,000 which is supposed to decrease to $44,000 annually, at an undetermined time
³ For the one RA which required O&M
⁴ Year 1 and year 2 O&M costs averaged,another source says after year 1 operating costs will be $110,000
⁵ Year 1 and year 2 O&M costs averaged.
⁶ $24,000 over 4 years.

2. Total Estimated Job Generation from RCRA Site Cleanup

Cost of Cleaning up RCRA Sites

Assuming that there are 4,700 RCRA TSD sites to be cleaned up, with an expenditure of at least $230 billion, then an estimated total of 1,080,979 job years are generated. Of this 928,370 job years are estimated for direct remedial action and 152,609 job years are estimated for O&M activities. This estimate is based on the following assumptions: that 15 percent of RA dollars, on average, are allocated to direct construction labor; that average hourly earnings are $18.40; and that 1500 hours constitute a full-time job for a year, while 2000 hours constitute a full-time job for a year of O&M work.

Total cleanup costs for RCRA sites is significantly higher than for NPL sites, with many estimates reaching $230 billion or more. While, in aggregate, assumed to be over $230 billion, the range of possible total costs from the University of Tennessee study was $130 billion to $450 billion, with a most appropriate range estimated at $203 billion to $265 billion (the average estimates for the least stringent and most stringent estimates).⁹⁷ But the total range, based on less stringent and more stringent requirements is very broad -- $105 billion to $600 billion. An estimate of environmental remediation costs in 1993⁹⁸ found that nearly one-third of the entire estimated cost for environmental remediation, a total of well over $450 billion, would be used to clean-up sites covered by RCRA. Cost estimates for RCRA sites may, however, be somewhat less accurate than estimates for NPL sites because most have not been well evaluated.⁹⁹

There are at least eight different categories of RCRA sites,¹⁰⁰ but over 40 percent of the burden fell to the cleanup of tanks and tank areas. Landfills, surface impoundments, and container areas accounted for 13 percent to 17 percent each of the total cost, and waste piles, land treatment units, and satellite areas accounted for 3 percent to 6 percent each of total RCRA cleanup cost.

Duration of Activities

The duration of a RCRA site cleanup is expected to be quite long, especially for O&M.¹⁰¹ Some cleanups, such as those requiring soil excavation and treatment, may have as little as two, four, or six years of O&M. Sites requiring pump and treat are more likely to require five, twenty, or one hundred years of O&M. The Cost of Remedial Action (CORA) Model developed for EPA by CH2M Hill in 1989, used a 60 year period for soil and RCRA caps, which included one replacement cap and 60 years of operations and maintenance (O&M).

3. Total Estimated Job Generation from DOE Site Cleanup

A total of 1,108,696 remedial action and O&M job years are projected to be generated for hazardous waste cleanup of the DOE nuclear weapons complex. This estimate is derived from the University of Tennessee's best estimate of $240 billion, using the assumptions that: remedial action costs are 73 percent of the total restorations costs and O&M costs are 16 percent of the total restoration costs; 15 percent of RA dollars, on average, are allocated to direct construction labor; average hourly earnings are $18.40; and 1500 hours constitute a full-time job for a year, while 2000 hours constitute a full-time job for a year of O&M work.

Total Estimated Job Generation from DOE UMTRA Sites

UMTRA represents a small portion of the cleanup work at DOE and the number of job years generated at UMTRA sites is only one percent of those generated at all DOE sites, with 11,183 remedial action jobs expected. (These projections assume that RA represents 73 percent of total cost, that direct construction labor accounts for 15 percent of RA. Average hourly earnings used are those generate by craft from the RRA data base and 1500 hours are assumed to constitute a full-time equivalent work year for a construction worker.)

The major operations and maintenance task at UMTRA sites is to treat contaminated groundwater. According to one expert,¹⁰² this is expected to last through 2014 (19 years) and cost $7.8 million per year, all 19 years, for a total cost of $148.2 million (six percent of the remediation cost) with treatment mostly through natural attenuation rather than pump and treat.

For remedial actions jobs by craft for all UMTRA sites see Table Y.

Cost of Cleaning Up DOE Sites

Funding for DOE cleanup programs is set at $12.3 billion for 1994 through 1998, with estimates of the total cleanup cost growing to the hundreds of billions of dollars. The University of Tennessee best estimate for cleanup of the DOE complex is $240 billion. DOE's mid-range estimate of its cleanup cost is $230 billion.¹⁰³

Cost of Cleaning Up DOE UMTRA Sites

Cost estimates for cleanup of UMTRA sites were also made by the General Accounting Office (GAO) in 1995.¹⁰⁴ GAO estimated the overall costs for the 23 UMTRA sites to be $2.315 billion, with at least 85 percent of those funds spent by year end 1995. The four UMTRA sites studied in this report range from the most expensive (Grand Junction) of the 23 sites to the 16th most expensive. Together the four sites account for 47 percent of total projected cleanup costs at the 23 UMTRA sites ($1.082 billion of $2.315 billion). The Grand Junction site alone accounts for 32 percent of the total UMTRA cost, at $746 million.

4. Total Estimated Job Generation from DOD Cleanup

Using the University of Tennessee's best estimate of the cost of DOD cleanup at $30 billion, 138,587 job years will be needed for remedial action and O&M activities over the next thirty years. This estimate is based on the assumptions that remedial action costs are 73 percent of the total restorations costs and O&M costs are 16 percent of the total restoration costs.

The Cost of Cleaning Up DOD Sites

The total estimated cost for all IRP activities over the next 20 years, according to EPA, is $24.5 billion.¹⁰⁵ The University of Tennessee estimated a total cost for cleanup of DOD sites at $30 billion and DOE estimated DOD cleanup costs at $25 billion. A 1991 EPA report estimated the RD/RA costs for DOD were $15.4 billion, of which $14 billion was to be for remedial action.¹⁰⁶ In addition, EPA experts estimated operations and maintenance (O&M) to cost $4 billion (consistent with DOE's estimate of DOD's 1992 O&M costs at $4.1 billion).

5. Total Estimated Job Generation from UST Cleanup

An estimated 309,511 job years are needed -- for both remedial actions and O&M activities -- to clean up petroleum and other hazardous wastes from the nation's underground storage tanks. Of this total 265,815 job years are for direct remedial action, and 43,696 job years are for O&M. This assumes that the total for UST cleanup is $67 billion,¹⁰⁷ that 73 percent of dollars allocated are for remedial action, that 15 percent of RA dollars are allocated to direct construction labor, that 1500 hours represent of a full time construction job, and that average hourly earnings are $18.40. The operations and maintenance estimate of 43,696 job years assumes that O&M costs are 16 percent of remediation costs ($67 billion), that construction labor costs are 15 percent of total RA costs, that average hourly earnings is $18.40 and 2000 hours represent a full-time job for a year.

Cost of Cleaning Up UST Sites

The University of Tennessee's best estimate for cleaning up UST sites is $67 billion. However, as the findings show, many factors need to be taken into consideration when developing that total cost estimate.

Number of Sites and Cost of UST Cleanup

There are several different estimates on the number of underground storage tanks existing and in need of cleanup:

• According to EPA, as of 1993, approximately 295,000 UST sites, containing at least 56 million cubic yards of contamination, required cleanup.¹⁰⁸ This includes 119,000 confirmed releases that have not been cleaned up yet, in addition to 176,000 projected releases. Previous studies have indicated that remedial costs per site can range from $2,000 to $400,000 with an average of $100,000.
• As of October 30, 1995 there were 1,093,105 active tanks registered and 983,877 closed tanks. There had been 303,635 confirmed releases, 239,671 cleanups initiated, 113,512 enforcement actions, and 8,600 emergency responses. Completed tank remediations numbered 90,529.¹⁰⁹
• When groundwater as well as soil are contaminated, there may be significant operations and maintenance costs. Those UST sites with groundwater contamination cost as much as 10 times more than tanks having only soil contamination ($125,000 to over $1 million vs. $10,000-$125,000).¹¹⁰

An estimated 138,587 job years are needed for both remedial and O&M activities to cleanup state and private sites. Of this, 119,022 job years are for direct remedial action, and 19,565 job years are O&M activities. This estimate is based on the assumptions that remedial action costs are 73 percent of the total restorations costs and O&M costs are 16 percent of the total restoration costs; that 15 percent of RA dollars, on average, are allocated to direct construction labor; that average hourly earnings are $18.40; and that 1500 hours constitute a full-time job for a year, while 2000 hours constitute a full-time job for a year of O&M work.

Of the sites identified by EPA, ASTSWMO, and Kensington; 2,844 had remedial construction completions (2689 State/Territory and 155 Federal) and 20,134 sites were still in the remedial process, with at least a preliminary assessment (11,000 State/Territory and 9,134 Federal).

If the sites completed or with major decisions about remediation are comparable to the universe of identified sites, then approximately 14 percent of the remedial actions and 36 percent of the removal sites had been completed by year end 1992.

Cost of Cleaning Up State and Private Sites

University of Tennessee study in 1991¹¹¹ estimated that each state or private sector site cleanup would cost an average of $1 million, for a total resource commitment of between $12 billion and $32 billion. Two separate analytic approaches were used by the University of Tennessee in making these estimates and "the two approaches tracked each other exceptionally well." The study estimated that through the year 2020 there would be 24,335 state sites requiring cleanup.¹¹² The University of Tennessee estimated that five to fifteen as many non-Federal program sites exist as are listed on the NPL and as are at Federal facilities.¹¹³

The data from approximately 3,500 sites provided costs for remediation of over $1.2 billion. (See Table Z.) The average site cleanup cost for state sites ranged from under $3,000 in Oklahoma to over $3 million in Colorado. For 18 of the 31 states that provided site data, the average costs were between $50,000 and $500,000. For PRP sites the average ranged from $1000 in Indiana to $7 million in Utah, with the largest number of estimates between $200,000 and $500,000.¹¹⁴

State sites usually cost less than federal sites to cleanup. For remedial actions studied by EPA,¹¹⁵ the average federal remedial construction completion cost $2.4 million whereas the average state completion cost $183,000. For completed removals, EPA reported an even greater difference between the average cleanup cost of a federal and a state site -- $7.1 million and $47,000, respectively.

Cost data are probably more reliable from some states than from others, due to higher levels of experience. New Jersey, for example, has completed 925 remedial actions and Massachusetts and Michigan more than 300 each, while 25 states reporting had less than 10 completions. (See Table V.)

NUMBER OF EMPLOYMENT EPISODES LIKELY TO BE HIGHER THAN ESTIMATED JOB YEARS

Because of peak work time at a site, at least three times as many individuals may perform work as estimated job years may predict. (See Table 14.) The demand for an individual's skills may be for only a few hours, days, or months based on subspecialty, task at hand, etc. These estimates do not include the need to cover absenteeism, turnover, and the need for individuals with highly specialized craft skills within a trade. On the other hand, because job projections are made in job years, there may be some offsetting reductions for individuals who work more than one full-time job year.

Peak time demand results primarily from the specialization of trades on a site and the need to hire specific individuals to perform specific tasks, and sometimes tasks of relatively brief duration. So, for example, at a specific site, like Bridgeport (BROS) in New Jersey, where averaged over the period of remedial action there was a requirement for an average of 59 workers a month, in the month of peak demand for workers, 203 were needed -- thus an overall site demand for 234 percent more individuals than the overall average would suggest. For plumbers and pipe fitters at BROS, however, the peak demand is 100 percent higher than average; for iron workers it was 900 percent higher; for electricians it was 450 percent higher; for laborers, operators, and drivers the demand was approximately double.

It is not enough to simply estimate demand for the amount of labor time on an environmental remediation job. One must focus on the total number of individuals who will be working on a site. This is not a simple arithmetic calculation:

• Simply Estimating Number of Hours To Be Worked May Not Be Accurate. Even if one could accurately estimate the number of labor hours to be worked on a site, using this as a basis for calculation would be misleading. Estimating employment episodes by estimating the number of hours of work and dividing by a full-time worker equivalent; say, 1500 hours for a construction worker, is a way to get a rough estimate, but this method is misleading, because remediation work is not spread evenly among crafts.
• Estimating Number of Hours By Craft May Not Be Accurate. While somewhat more accurate for estimating training needs, estimating the number of hours of work by craft divided by a full-time equivalent number of hours is also not enough, because work is not spread evenly over time, especially not by craft. For example, if there is a large decontamination pad to be built, a site is likely to employ a significantly larger number of iron workers and carpenters than another site. At some sites there is a more prevalent share of heavy equipment operators. At others there is a high percentage of drillers or electricians or laborers. Labor mix depends on the type of remediation, management decisions, and a number of other factors.
• Estimating Number of Hours By Craft By Month Makes Estimates More Realistic. More accurate than an assessment of overall labor requirements or overall labor requirements by craft, is to determine the peak month of work at a site for each craft and determine what number of individuals, at a minimum, would need to be employed to meet those peak month demands. At one site, Rifle, the peak month required more than 6 times the number of workers than the average month. (See Table 14.) At Bog Creek, Bonfouca, and New Lyme more than three times the number of workers were needed. On average for the sixteen sites, 332 percent more employment episodes were needed, based on peak month estimates, than would have been expected from average month labor needs.

Categories:

For some trades, the increased number of employment episodes needed, when estimated by peak month, are especially large. At the two sites, Bonfouca and Lipari, where bricklayers made up one percent or more of total hours, the increase was 300 percent and 500 percent respectively. For drivers the increases were as high as 620 percent at Rifle; for laborers at Rifle the increase was almost 800 percent; for iron workers, at BROS, the increase was 900 percent. For plumbers and pipe fitters at Bros the increase was 1100 percent and at Bog Creek 550 percent.

• Estimating Number of Hours By Specialty Within Each Craft, By Month, Would Make Estimates Even More Realistic. These data, unfortunately were not available for this analysis. But it is important to note that many jobs at hazardous waste sites are very specialized. They may require specialized workers to operate a specific type of equipment, and so the number of operator hours in a given month might involve multiple operators even though the total number of hours is less than full-time.

At peak demand, from the year 2000 through 2005, NPL site cleanup alone is likely to generate over 200,000 job years, requiring the employment episodes for as many as 600,000, maybe more, HAZMAT workers.

JOB PROJECTIONS OVER TIME

The arduous task of cleaning up the nation's hazardous waste sites has begun. At what pace it will continue depends on public policy and budget decisions. The projections in this section are based, not on University of Tennessee estimates, but rather on a 1993 study done for DOE,¹¹⁶ which compiled work plans over time across regulatory bodies and legislative jurisdiction. (See Tables 8 and 15 and Chart 3.) As a result, the number of job years projected in this section of the report does not tightly parallel discussion of job years in other sections. For example, the DOE study projected total cleanup costs over 30 years of $983 billion, higher than the $752 billion estimates of the University of Tennessee. For the 20 years, 1990-2010, for which projections are made in this section, the estimated cleanup costs by DOE are $758 billion -- with higher projected proportionate costs for RCRA and State/Private sites than the UT study and lower proportionate costs for cleaning up DOE sites.

Using the same assumptions used throughout this report to make projections, remediation job demand is expected to grow by 60 percent, or almost 300,000 jobs, from the 1990-1995 five year period through the five year period 1995-2000 -- from 447,000 to 740,000. (See Table 16 and Chart 4.) Demand for jobs continues to grow by nearly another 300,000 in the 2000-2005 time interval. During this peak period nearly 2 million jobs will require workers. As many as 7.5 million more workers will require training -- either basic or refresher. Demand remains high from 2005-2010 and then begins to taper off -- with a rather optimistic assumption that most cleanup activities will be completed in 25 to 30 years.

Table 16

Andelman Julian and Dwight Underbill, Health Effects From Hazardous Waste Sites, Chelsea, Lewis, 1987.

Apogee Research, Inc., A Report on Clean Water Investment and Job Creation, March 20,1992.

The Atlanta Constitution, "Hot Air From the Business Roundtable," Monday, March 14, 1988.

Bechtel, Annual Report, December 1992.

Been, Vacua, "Unpopular Neighbors: Are Dumps and Landfills Sated Equitably?" Resources, Spring 1994, No. 115.

Bueckman, Donna, Camber Sunita, and Milton Russell, Underground Storage Tanks: Resource Requirements For Corrective Action, Waste Management Research and Education Institute, University of Tennessee, Knoxville, December 1991.

Bullard, Robert D., People of Color: Environmental Groups, Environmental Justice Resource Center, Clark Atlanta University, Atlanta, 1994.

California (State of) Governor's Office of Emergency Services, Hazardous Materials Unit, Hazardous Materials Incidents: California, January 1990-December 1991, April 1994.

Canada Employment and Immigration Advisory Council and Management Information Services, Inc., Environment and Employment in Canada: The Final Report of an Interactive Symposium for Labor, Business and Environmentalists, MISI, Washington, DC, April 1992.

Chakravarty, Subrata N, "Dean Buntrock's Green Machine," Forces, August 2, 1993.

Chemical Manufacturers Association, "Impact Analysis of RCRA Corrective Action and CERCLA Remediation Programs," Washington, DC, CMA, 1988, cited in Shreekant Gupta, George Van Houtven, and Maureen Cropper, "Do Benefits and Costs Matter in Environmental Regulation? An Analysis of EPA Decisions under Superfund."

Chemical Regulation Reporter, "Current Report Index," Vol. 17, No.3, Bureau of National Affairs, Washington, DC, April 16, 1993.

Chicago Tribune, "Study Shows Going Green Helps Economy Grow Too," March 3, 1993.

Clean Sites, Improving Remedy Selection: An Explicit and Interactive Process for the Superfund Program, October 1990.

Clean Sites, Lone Pine Landfill Site: A Description of Site Activities, May 1992.

Colglazier, E.W., T. Cox, and K. Davis, Estimation Of Resource Requirements for NPL Sites, Waste Management Research and Education Institute, University of Tennessee, Knoxville, December 1991.

Colglazier, E.W., Mary English, and Milton Russell, Hazardous Waste Remediation: The Task Ahead, Waste Management Research and Education Institute, The University of Tennessee, Knoxville, December 1991.

Committee for the National Institute for the Environment Congressional Research Service Report for Congress, "Superfund Reauthorization issues in the 104th Congress," Washington, DC, January 3, 1996.

Day, Sue, Andrew Whitford and Eiman Zeinelabdin, State and Private Sector Cleanups, Waste Management Research and Education Institute, University of Tennessee, Knoxville, December 1991.

Delp, Linda, A Cross-Sectional Survey of Workers and Their Training Needs at 29 Hazardous Waste Sites, University of California, Los Angeles, December 1993.

Doty, Carolyn, Amelia Crotwell, and Curtis Travis, Cost Growth for Treatment Technologies at NPL Sites and The Effectiveness of Groundwater Pumping As A Restoration Technology, Waste Management Research and Education Institute, University of Tennessee, Knoxville, December 1991.

Durett, Dan, Environmental Justice: Breaking New Ground, 3rd Edition, Committee for the National Institute for the Environment, Washington, DC, October 26, 1994.

Eastern Research Group, Inc., Preparation of Data to Support a Regulatory Analysis and Environmental Impact Assessment of the Proposed Standard for Working at Hazardous Waste Sites, Arlington, MA, April 1987.

Eastern Research Group, Inc., Survey of Emergency Response Training Programs, Arlington, MA, Completion of Requirements for Task Order No.3 Option Year 2, Contract No.J-9-F-8-0035, Submitted to OSHA.

Ehrbar, Robert, The Lone Pine Landfill Superfund Project Freehold, New Jersey, December 1993.

The Energy Daily, "Paper Chase," Vol. 15, No. 53, March 23, 1987.

Environment Reporter, "$170 Billion Spent on Environment in 1992; Market Continuing to Grow, Economists Says," Vol. 23, No. 47, Bureau of National Affairs, Inc, Washington, DC, March 19, 1993.

Environment Reporter, "Ten-Month Index-Summary," Vol.23, No.46, Bureau of National Affairs, Inc., Washington, DC, March 12, 1993.

Executive Order of the President 12898, "Federal Actions to Address Environmental Justice In Minority Populations and Low Income Populations", White House, Washington, DC, February 11, 1994.

Federal Emergency Management Agency, Digest of Federal Training in Hazardous Materials, FEMA, Washington, DC, July 1987.

Fernald Environmental Restoration Management Company, "Fernald Environmental Management Project: Fiscal Year 1994 Roadmap," prepared for the Department of Energy, Ohio, September 30, 1993.

Finn, M.A., C.B. Holmes, S.A. Hunt, and R.R. Lewis, Environmental Restoration and Waste Management Manpower Needs Assessment: U.S. Department of Energy Complex, Pacific Northwest Laboratories, June 1992.

Goeltz, Richard, Ho-Ling Hwang, Milton Russell, Bruce Tonn, and John Warren, Costs of RCRA Corrective Action Interim Report, Waste Management Research and Education Institute, University of Tennessee, Knoxville, December 1991.

Harden, Blaine, "Nuclear Reactions," Washington Post Magazine, May 5, 1996, p.14.

Houghton, Aimee, and Lenny Siegel, Military Contamination and Cleanup Atlas for the United States - 1995, The Pacific Studies Center and CAREER/PRO, San Francisco State University, San Francisco, CA, 1995, p. 2.

Hughes, Colleen, Ralph Kummler, Robert Powitz and Catherine Witt, Evaluating Hazardous Waste Education and Training, Wayne State University, Detroit, November 30, 1990.

Illinois Institute of Technology/IIT Research Institute, Coalition on Superfund Research Report: Consolidated Research Report, Center for Hazardous Waste Management, IIT, Chicago, September 1989.

Inside EPA, "An Exclusive Report on the United States Environmental Protection Agency," Weekly Report, Washington, DC, March 12, 1993.

International Association of Environmental Testing Laboratories, An Analysis of the Market and Business Opportunities in the Environmental Testing Laboratory Industry, October 1991.

International Association of Fire Fighters, "Alternative for Baseline Provision of Fire Protection and Emergency Medical Delivery for Texas City, Texas: A Rebuttal to CRESAP Report."

International Association of Fire Fighters, Safe Fire Fighter Staffing: Critical Considerations, IAFF, Washington, DC, 1993.

Irene Hall, Ramana Dhara, Wendy Kaye, Patricia Price-Green, "Surveillance of Hazardous Substance Releases and Related Health Effects," Archives of Environmental Health, Vol.49, (No.1), January/February 1994, pp. 45-48.

Johnson, Barry, Assistant Surgeon General, Assistant Administrator, ATSDR, PHS, USDHHS, "Testimony Before the Subcommittee on Superfund, Recycling, and Solid Waste, Management," U.S. Senate, May 6, 1993.

Katherine Probst, Don Fullerton, Robert Litan, and Paul Portney, Footing the Bill for Superfund Cleanups, The Brookings Institution and Resources for the Future, Washington, DC, 1995.

Kensington Systems, Inc., "State and Territory System Documentation," for the U.S. Environmental Protection Agency, Washington, DC, August 1994.

Kopp, Raymond and Kerry Smith, Valuing Natural Assets, Resources for the Future, Washington, DC, 1993.

LaGuardia, Thomas and William Manion, Decommissioning Handbook, Nuclear Energy Services, Inc., Prepared for DOE, DOE/EV/10128-1, Danbury, CT, November 1980.

Lawrence Livermore National Laboratory, University of California, Jobs Needs Analysis Volume I: Division Staff and Support Personnel, LLNL, Environmental Protection Department, Hazardous Waste Management Division, June 1991.

Lenny Siegel, "Jobs From a Typical Cleanup Task: Underground Storage Tank Removal," Pacific Studies Center, Mountain View , CA, September 1994.

Levine Associates, Evaluating the Environmental Health Work Force, January 1988.

Lippitt, John, William Martin and Timothy Prothero, Hazardous Waste Handbook for Health and Safety, Stoneham, Butterworth, 1987.

Logistics Management Institute, Environmental Restoration: Remedial Action Work Breakdown Structure, IR518LN1, McLean, VA, December 1994.

Management Information Services, Inc., The Net Impact of Environmental Protection on Jobs and the Economy, Washington DC, March 1993.

Management Information Services, Inc., News Release, "1992 United States Environmental Spending Stimulates Economy, Creating $170 Billion in Revenues & 4 Million Jobs," Washington, DC, March 8, 1993.

Meyer, Stephen M., Environmentalism and Economic Prosperity: Testing the Environmental Impact Hypothesis, Project on Environmental Politics an Policy, Massachusetts Institute of Technology, Cambridge, MA, October 5, 1992.

National Constructors Association, et. al., Hazardous Waste Cleanup Opportunities & Issues, Washington, D.C., September 16, 1986.

National Defense Authorization Act, Fiscal Year 1989 (p.L. 100-456-Sept, 29, 1988), as amended by National Defense Authorization Act, Fiscal Year 1991 (p.L. 101-510-Nov. 5, 1990), and National Defense Authorization Act, Fiscal Year 1992 and 1993 (p.L. 102-190-Dec. 5, 1991).

National Environmental Education and Training Center, "The State of Hazardous Materials Education, Training, and Research," Prepared by Ellen Warhit, NEETC, Indiana, PA, May 1995.

National Environmental Education and Training Center, "The U.S. Environmental Remediation Market," Prepared by Ellen Warhit, NEETC, Indiana, PA, June 1995.

National Fire Protection Association, Fire Protection Handbook, Fifteenth Edition, Quincy MA.

National Public Radio, "Segment #11: The Failure of the Superfund Law - Part I," Morning Edition, September 12, 1994.

Nordheimer, Jon, "Residue of Fear Remains as Toxins are Removed," New York Times, November 6, 1994.

Nuclear Energy Services, Inc. Automation Industries, Decommissioning Handbook, Prepared for U.S. Department of Energy, DOE/EJ/101/28-1,Nov, 1980.

OCAW and Greenpeace, "Greenpeace, OCAW Take on Jobs vs. Environment Conflict. Call for Clean Energy Transition with Worker Protection," November 23, 1993.

President's Council on sustainable Development, Sustainable America, U.S. Government Printing Office, Washington, DC, March 1996.

Reimherr, Joyce, Information Access: Its Role in Learning and Reinforcement of Worker Health and Safety Training for Hazardous Waste Operations and Emergency Response, National Clearinghouse for Worker Safety and Health Training for Hazardous Materials, Waste Operations, and Emergency Response, Silver Spring, MD, April 1991.

Renner, Michael G., "Jobs in a Sustainable Economy," World Watch, Paper 104, September 1991.

Renner, Michael G., "Saving the Earth, Creating Jobs". World Watch, January-February 1992.

Resources for the Future, "RFF Database of Superfund NPL Sites," from World Wide Web, 1995.

Revensz, Richard and Richard Stewart, Editors, Analyzing Superfund: Economics, Science, and Law, Resources for the Future, Washington, DC, 1995.

Robinson, GAIUS, Hazardous Chemical Spill Cleanup, Nodes Data, Park Ridge, II, 1979.

R.S. Means Company, Inc., ECHOS 1996 Environmental Restoration Cost Books Assemblies Costs/Unit Costs., Kingston, MA, 1996.

Russell, Milton, Assessment of Incineration As A Treatment Method for Liquid Organic Hazardous Materials, March 1985.

Rusell Milton, Project Director, The Superfund Process: Site Level Experience, Waste Management Research and Education Institute, University of Tennessee, Knoxville, December 1991.

San Francisco Urban Institute, San Francisco State University, Career/Projections, Vol 1, Issue 1, October 30, 1993.

Scruggs, Yvonne, "Environmental Economic Equity: The Potential for Stimulating Job Training and Relieving Minority Unemployment Through Hazardous Waste Clean-Up Operations", Unpublished Document, June 28, 1994.

Seigel, Lagneia, et al., The U.S. Military's Toxic Legacy: America's Worst Environmental Enemy, National Toxic Campaign Fund.

Select Update, "The Growth Outlook for Environmental Services in the 1990s," First Quarter 1991.

Seminario, Margaret, "The Training of Targeted Groups," Toxicology and Industrial Health, Vol. 5, No. 4, 1989, pp.85-128.

Singh, Vern, P.E., and James Lanzo, P.E., "Lessons Learned from Remedial Design of the Helen Kramer Landfill Superfund Site," URS Consultants, Inc., Buffalo, New York.

Smith, Michael J., et. al. "Characteristics of Successful Safety Programs," Journal of Safety Research; Vol.10, No.1:, Spring 1978, pp.5-15.

Stanton, Michael. "Hazardous Wastes: Who's Cleaning Up?", Occupational Outlook Quarterly, Winter 1987, pp. 3-6.

Tonn, Bruce, Russell Milton, Hwang Ho-Ling, Goeltz Richard, and Warren John, Cost of RCRA Corrective Action: Interim Report, Waste Management Research and Education Institute, University of Tennessee, Knoxville, December 1991.

Turner, Keith A., Potential for Future Shortages of Technical Personnel for National Clean-Up of Uncontrolled Hazardous Waste Sites, Colorado School of Mines, Department of Geology, Golden, CO, November 30, 1984.

United States Army Corps of Engineers, Hazardous and Toxic Waste Contracting Problems, Fort Belvoir, VA, July 1990.

United States Congress, Congressional Budget Office, Analyzing The Duration Of Cleanup At Sites On Superfund's National Priorities List, Washington, DC, March 1994.

United States Congress, Congressional Budget Office, Prepared by Bender, Peter, The Total Costs of Cleaning Up Nonfederal Superfund Sites, U.S. Government Printing Office, Washington, DC, January 1994.

United States Congress, Office of Technology Assessment, Hazards Ahead: Managing Cleanup Worker Health and Safety at the Nuclear Weapons Complex, OTA-BP-0-85, U.S. Government Printing Office, Washington, DC, March 1993.

United States Department of Commerce, National Technical Information Service, Environmental Restoration and Waste Management Manpower Needs Assessment: United States Department of Energy Complex, Springfield, VA, June 1992.

United States Department of Defense, Defense Environmental Restoration Program Annual Report to Congress for Fiscal Year 1992, Washington, DC, April 1993.

United States Department of Education, Business and Education Standards Programs, National Voluntary Skills Standard, Washington, DC, January 1995

United States Department of Energy (EM-24), The Demand For Environmental Restoration Services, Prepared by Independent Project Analysis, Inc., 93-DOE-007, Washington, DC, May 1993.

United States Department of Energy, Environmental Restoration and Waste Management Five-Year Plan, Fiscal Years 1994-1998, Vol.1 and Installation Summary: Washington, DC, January 1993.

United States Department of Energy, Environment Safety and Health Progress Assessment of the Savannah River Site, Washington, DC, March 1993.

United States Department of Energy, Estimating the Cold War Mortgage, The 1995 Baseline Environmental Report, Executive Summary, Volume I, and Volume II , DOE/EM-0232, March 1995.

United States Department of Energy, Hanford Environmental Restoration Project Plan, November 1992.

United States Department of Energy, Manpower Assessment Brief, Number 26, Washington, DC, August 1994.

United States Department of Energy, "National Defense Authorization Act, Fiscal Year 1989 (P.L. 100-456-Sept. 29, 1988)," as amended by National Defense Authorization Act, Fiscal Year 1991 (P.L. 101-510-Nov. 5, 1990), and National Defense Authorization Act, Fiscal Years 1992 and 1993 (P.L. 102-190-Dec. 5, 1991).

United States Department of Energy, Fernald Field Office, Operable Unit 3: Record of Decision for Interim Remedial Action, Fernald Environmental Management Project, Fernald, Ohio, May 1994.

United States Department of Energy, Office of Environmental Restoration and Waste Management, Project Performance Study, Prepared by Independent Project Analysis, Inc., Washington, DC, November 1993.

United States Department of Energy, Chicago Operations Office, Roadmap Document, prepared for Environmental Restoration and Waste Management Program, Chicago, September 1993.

United States Department of Health & Human Services, Evaluating the Environmental Health Work Force, Prepared by Levine Associates, Inc., Rockville, MD, January 1988.

United States Department of Health & Human Services, Health Resources and Services Administration, Bureau of Health Professions and the Agency for Toxic Substances and Disease Registry, Division of Health Education, Evaluating Hazardous Waste Education & Training, Rockville, MD, November 30, 1990.

United States Department of Health & Human Services, Public Health Service, Agency for Toxic Substances and Disease Registry, Hazardous Substances Emergency Events Surveillance, Annual Report, Atlanta, 1991.

United States Department of Health & Human Services, Agency for Toxic Substances and Disease Registry, Public Health Service, Hazardous Substances Emergency Events Surveillance (HSEES), Annual Report, Atlanta, 1992.

United States Department of Health & Human Services Occupational Safety and Health Guidance Manual for Hazardous Waste Site Activities, Cincinnati, OH, 1987.

United States Department of Labor, Employment and Training Administration, Environmental Protection Careers Guidebook, Washington, DC, 1980.

United States Department of Labor, OSHA, Office of Information,"OSHA Releases Hazardous Waste Incinerators Safety Study Results," News, USDL #91-247, Washington, DC, May 23, 1991.

United States Department of Labor, OSHA, Office of Regulatory Analysis, Regulatory Impact and Regulatory Flexibility Analysis of the Occupational Safety and Health Standard for Hazardous Waste Operations and Emergency Response (29 CFR Part 1910), Washington, DC, December 14, 1988.

United States Environmental Protection Agency, Abstract of Remediation Case Studies, EPA-542-R-95-001, Prepared by the Member Agencies of the Federal Remediation Technologies Round Table, Washington, DC, March 1995.

United States Environmental Protection Agency, An Assessment of Superfund Benefits in Selected Geographic Areas, Washington, DC, October 1995.

United States Environmental Protection Agency, Experience Curves, Innovative and Diffusion: Application to RCRA Corrective Action, Draft Report, Prepared for the Office of Solid Waste, Prepared by Sobotka & Company, Inc., Washington, DC, April 1992.

United States Environmental Protection Agency, Guide to Documenting Cost and Performance for Remediation Projects, EPA-542-B-95-002, Prepared by the Member Agencies of the Federal Remediation Technologies Round table, Washington, DC, March 1995.

United States Environmental Protection Agency, Hazardous and Toxic Waste Contracting Problems, Washington, DC, July 1990.

United States Environmental Protection Agency, Innovative Treatment Technologies: Annual Status Report (Seventh Edition) Applications of New Technologies at Hazardous Waste Sites, EPA-542-R-95-008 Number 7, Revised, Washington, DC, September 1995.

United States Environmental Protection Agency, "National Priorities List for Uncontrolled Hazardous Waste Sites; Final Rule," Federal Register, December 16, 1994.

United States Environmental Protection Agency, "National Priorities List for Uncontrolled Hazardous Waste Sites, Proposed Rule No.16," Federal Register, Volume 59, No. 103, Washington, DC, May 31, 1994.

United States Environmental Protection Agency, The New Superfund, What Is It, How It Works, Washington, DC, August 1987.

United States Environmental Protection Agency, Progress Toward Implementing Superfund Fiscal Year 1990, Report to Congress, EPA/540/8-91/004, Washington, DC, February 1992.

United States Environmental Protection Agency, Remediation Case Studies: Soil Vapor Extract, EPA-542-R-95-004, Prepared by the Member Agencies of the Federal Remediation Technologies Round Table, Washington, DC, March 1995.

United States Environmental Protection Agency (Elliott Laws, Assistant Administrator). "Responses to 21 Questions" from Hon. John Dingell, U.S. House of Representatives to EPA Administrator Carol Browner, January 28, 1994,

United States Environmental Protection Agency, Superfund Administrative Improvements, Washington, DC, June 23, 1993.

United States Environmental Protection Agency, Superfund Emergency Response Actions: A Summary of Federally Funded Removals, Fourth Annual Report - Fiscal Year 1989, EPA/540/8-90/014, September 1990.

United States Environmental Protection Agency, with the Association of State and Territorial Solid Waste Management Officials, Underground Storage Tank and Leaking Underground Storage Tank Studied, Washington, DC, November 1995.

United States Environmental Protection Agency, Office of Emergency and Remedial Response, Interagency Agreement Between the Corps of Engineers and EPA in Executing P.L. 96.510 (CERCLA), OSWER Directive 92952-3, Washington, DC, December 3, 1984.

United States Environmental Protection Agency, Office of Emergency and Remedial Response, Joint Corps/EPA Guidance, Washington, DC, June 24, 1983.

United States Environmental Protection Agency, Office of Emergency and Remedial Response, National Priority List, Supplemental List and Supporting Materials, Washington, DC, July 1989.

United States Environmental Protection Agency, Office of Emergency and Remedial Response Superfund Risk Assessment Information Directory, Washington, DC, November 1986.

United States Environmental Protection Agency, Office of Emergency & Remedial Response Hazardous Site Control Division, An Analysis of State Superfund Programs: 50-State Study, Washington, DC, September 1989

United States Environmental Protection Agency, Office of Emergency and Remedial Response, Office of Program Management, Policy and Contracts Assessment Staff, DRAFT Economic Impact of the Superfund Program: Fiscal Years 1981 - 1992, Washington, DC, July 1994.

United States Environmental Protection Agency, Office of Emergency and Remedial Response, Office of Program Management, Policy and Contracts Assessment Staff, DRAFT Technical Report, Economic Impact of the Superfund Program: 1981 - 1992, Washington, DC, June 16, 1994.

United States Environmental Protection Agency, Office of Emergency and Remedial Response, Office of Solid Waste and Emergency Response, Environmental Response Division, Washington, DC., Removal Cost Management System: User's Guide.

United States Environmental Protection Agency, Office of Health Research, et. al., Inventory of Exposure-Related Data Systems Sponsored by Federal Agencies, Washington, DC, May 1992.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, Abstract Proceedings "Fifth Forum on Innovative Hazardous Waste Treatment Technologies: Domestic and International," EPA/540IR/94/503, Chicago, Illinois, May 3-5, 1994.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, Background Information: National Priorities List, Final Rule, Publication 9320.7-041, Washington, DC, May 1994.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, Cleaning Up the Nation's Waste Site: Markets and Technology Trends, EPA/542-R-92-012, Washington, DC, April 1993.

United States Environmental Protection Agency: Office of Solid Waste and Emergency Response, Controlling UST Cleanup Costs, Fact Sheet 1: Hiring a Contractor, Washington, DC, May 1992.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, Making It Work: The Fire Service and SARA Title III, EPA-550-K-93-004, Series 9, No. 5, Washington, DC, October 1993.

United States Environmental Protection Agency: Office of Solid Waste and Emergency Response, Memo from Mr. Henry Stevens, Site Assessment Branch, to Mr. Robert D. Putnam, Putnam Environmental Services, Washington, DC, November 2, 1989.

United States Environmental Protection Agency, Office of Solid Waste & Emergency Response, The Nation's Hazardous Waste Management Program at a Crossroads, The RCRA Implementation Study, EPA/530-SW-90-069, Washington, DC, July 1990.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, OSWER Environmental Justice Task Force Draft Final Report, OSWER 9200.3-16, Washington, DC, April 25, 1994.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, An Overview of the Outyear Liability Model (OLM), Draft, Publication: 9200.2-15FS, Washington, DC, January 1993.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, An Overview of Underground Storage Tank Remediation Options, EPA 510-F-93-029, October 1993.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, Presumptive Remedy for CERCLA Municipal Landfill Sites, Directive No. 9355.0-49FS, Washington, DC, September 1993.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, Presumptive Remedies: Policy and Procedures, Directive: 9355.0-47FS, Washington, DC, September 1993.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, Presumptive Remedies: Site Characterization and Technology Selection For CERCLA Sites With Volatile Organic Compounds In Soils, Directive: 9355.0-48FS, Washington, DC, September 1993.

United States Environmental Protection Agency, Office of Solid Waste and Emergency Response, A Report on State/Territory Non-NPL Hazardous Waste Site Cleanup Efforts for the Period 1980 - 1992, OSWER 9242.2-09, Washington, DC, July 1994.

United States Environmental Protection Agency: Office of Solid Waste and Emergency Response, Office of Policy, Planning, and Information, 1985 National Biennial Report of Hazardous Waste Generators and Treatment, Storage and Disposal Facilities Regulated Under RCRA, Volume 1: Summary, Washington, DC, March 1989.

United States Environmental Protection Agency, Office of Solid Waste, Supplemental Information on Corrective Action Management Units ("CAMUS"), (CASA 50001), Washington, DC, October 16, 1992.

United States Environmental Protection Agency, Office of Research and Development, Survey of Materials-Handling Technologies Used at Hazardous Waste Sites, EPA/540/2-91/010, Washington, DC, June 1991.

Environmental Protection Agency, Office of Underground Storage Tanks. Data Sheets. October 30, 1995.

United States Environmental Protection Agency, Task Force on Technology Cooperation and Export Assistance, Global Markets for Environmental Technologies: Defining a More Active Role for EPA Within a Broader United States Government Strategy, Washington, DC, November 12, 1992.

United States General Accounting Office, DOE Management: Impediments to Environmental Restoration Management Contracting, GAO/RCED-92-244, Washington, DC, August 1992.

United States General Accounting Office, FEDERAL FACILITIES: Agencies Slow to Define the Scope and Cost of Hazardous Waste Site Cleanups, GAO/RCED-94-73, Washington, DC, April 1994.

United States General Accounting Office, Testimony: "GAO's Views on DOE's Environmental Restoration and Waste Management Five-Year Plan," GAO/T-RCED-90-16, Statement of J. Dexter Peach, Assistant Comptroller General Resources, Community, and Economic Development Division before the Committee on Governmental Affairs, United States Senate, Washington, DC, November 14, 1989.

United States General Accounting Office, HAZARDOUS WASTE: Corrective Action Cleanups Will Take Years to Complete, GAO/RCED-88-48, Washington, DC, December 1987.

United States General Accounting Office, HAZARDOUS WASTE: DOD Estimates for Cleaning Up Contaminated Sites Improved but Still Constrained, GAO/NSIAD-92-37, Washington, DC, October 1991.

United States General Accounting Office, HAZARDOUS WASTE: EPA Cleanup Requirements-DOD Versus Private Entities, GAO/NSIAD-89-144, Washington, DC, July 1989.

United States General Accounting Office, HAZARDOUS WASTE: Much Work Remains to Accelerate Facility Cleanups, GAO/RCED-93-15, Washington, DC, January 1993.

U.S. General Accounting Office, HAZARDOUS WASTE: Status and Resources of EPA's Corrective Action Program, GAO/RCED-909-144, Washington, DC, April 1990, p.1.

United States General Accounting Office, HAZARDOUS WASTE: Tinker Air Force Base is Making Progress in Cleaning Up Abandoned Sites, GAO/NFIAD-87-164-BR, Washington, DC, July 10, 1987.

United States General Accounting Office, HAZARDOUS WASTE SITES: State Cleanup Status and Its Implications for Federal Policy, GAO/RCED-89-164, Washington, DC, August 1989.

United States General Accounting Office, High Risk Series, Superfund Program Management, GAO/HR-93-10, Washington, DC, December 1992.

United States General Accounting Office, MILITARY ORDNANCE: Cleanup Activities at the Former Raritan Arsenal, Washington, DC, August 1, 1992.

United States General Accounting Office, NUCLEAR CLEANUP: Difficulties in Coordinating Activities Under Two Environmental Laws, GAO/RCED-95-66, Washington, DC, December 1994.

United States General Accounting Office, NUCLEAR FACILITY CLEANUP: Centralized Contracting of Laboratory Analysis Would Produce Budgetary Savings, GAO/RCED-94-118, Washington, DC, May 1995.

United States General Accounting Office, NUCLEAR WASTE: Further Improvement Needed in the Hanford Tank Farm Maintenance Program, GAO/RCED-95-29, Washington, DC, November 1994.

United States General Accounting Office, RADIOACTIVE WASTE: Status of Commercial Low-Level Waste Facilities, GAO/RCED-95-67, Washington, DC, May 1995.

United States General Accounting Office, SUPERFUND: Civilian Federal Agencies Slow to Clean Up Hazardous Waste, GAO/RCED-87-153, Washington, DC, July 1987.

United States General Accounting Office, SUPERFUND: EPA Cost Estimates Are Not Reliable or Timely, GAO/AFMD-92-0, Washington, DC, July 1992.

United States General Accounting Office, SUPERFUND: Estimates of Number of Future Sites Vary, GAO/RCED-95-18, Washington, DC, November 1994.

United States General Accounting Office, Peach, Dexter, SUPERFUND: Extent of the Nation's Potential Hazardous Waste Problem Still Unknown, GAO/RCED-88-44, Washington DC, December 1987.

United States General Accounting Office, Testimony of Peter F. Guerrero, Associate Director, Environmental Protection Issues Resources, Community, and Economic Development Division, "Potential Liability of Property/Casualty Insurers for Costs of Cleaning Up Hazardous Waste Sites," 049584/142327, before the Subcommittee on Policy Research and Insurance Committee on Banking, Finance and Urban Affairs, House of Representatives, Washington, DC, September 27, 1990.

United States General Accounting Office, URANIUM MILL TAILINGS: Cleanup Continues, but Future Costs Are Uncertain, GAO/RCED-96-37, Washington, DC, December 1995, pp.25-26.

United States Senate, Committee on Environment and Public Works, The Comprehensive Environmental Response, Compensation, and Liability Act of 1980 (Superfund) (P.C. 96-510), Senate Committee on Environment and Public Works, Washington, DC, 1987.

United States Senate, Committee on Environment and Public Works, The Solid Waste Disposal Act, EPA/530-SW-85-022, U.S. Government Printing Office, Washington, DC, 1987.

Washington Post, "Caution: HAZMATS Ahead," Health Section, September 20 1994.

Washington Post, "Environmental Entrepreneurs Cleaning Up," October 22, 1990.

Wells, Barbara B., "Restrictions Imposed on Contaminated Sites: A Status of State Actions", National Governors' Association: Environment, Health and Safety Program, National Resources Policy Studies Unit, Center for Policy Research.