Clean Water: Better Information and Targeted Prevention Efforts  
Could Enhance Spill Management in the St. Clair-Detroit River	 
Corridor (07-JUN-06, GAO-06-639).				 
                                                                 
Spills of oil and hazardous substances in the St. Clair-Detroit  
River corridor have degraded this border area between the United 
States and Canada and are a potential threat to local drinking	 
water supplies. Within the United States such spills are reported
to the National Response Center (NRC), and in Canada to the	 
Ontario Spills Action Centre. This report discusses (1) how many 
oil and hazardous substance spills greater than 50 gallons (or of
an unknown volume) were reported in the corridor from 1994 to	 
2004, and how accurately reported spills reflect the extent of	 
actual spills; (2) what processes are used to notify parties of  
spills, and if they contain explicit requirements for reporting  
times and spill magnitude; and (3) the extent of Environmental	 
Protection Agency (EPA) and the Coast Guard's spill prevention	 
efforts and enforcement activities in the corridor from 1994 to  
2004.								 
-------------------------Indexing Terms------------------------- 
REPORTNUM:   GAO-06-639 					        
    ACCNO:   A55219						        
  TITLE:     Clean Water: Better Information and Targeted Prevention  
Efforts Could Enhance Spill Management in the St. Clair-Detroit  
River Corridor							 
     DATE:   06/07/2006 
  SUBJECT:   Environmental monitoring				 
	     Hazardous substances				 
	     Oil spills 					 
	     Potable water					 
	     Water pollution					 
	     Water pollution control				 
	     Water quality					 
	     Data collection					 
	     Data integrity					 
	     International cooperation				 
	     Canada						 
	     EPA Spill Prevention Control and			 
	     Countermeasures Plan				 
                                                                 
	     National Pollution Discharge Elimination		 
	     System Program					 
                                                                 
	     St. Clair-Detroit River Corridor			 
	     Superfund Program					 

******************************************************************
** This file contains an ASCII representation of the text of a  **
** GAO Product.                                                 **
**                                                              **
** No attempt has been made to display graphic images, although **
** figure captions are reproduced.  Tables are included, but    **
** may not resemble those in the printed version.               **
**                                                              **
** Please see the PDF (Portable Document Format) file, when     **
** available, for a complete electronic file of the printed     **
** document's contents.                                         **
**                                                              **
******************************************************************
GAO-06-639

     

     * Results in Brief
     * Background
     * Agency Spill Data Are Not Sufficient to Accurately Determine
          * Agency Officials Believe Responsible Parties May Not Report
          * NRC Data Contains Multiple Reports of the Same Spills
          * EPA Does Not Eliminate All Duplicate Spill Reports or Update
          * The Coast Guard Updates Most Spill Data but Does Not Update
          * Number of Reported CSOs and Industrial Permit Violations Exc
     * Spill Notification Occurs between and among Many Different P
          * Spill Notification Is Multi-Faceted and May Follow Many Diff
          * Two Agreements Outline Notification between the United State
          * Neither Agreement is Explicit about When or How Quickly Noti
          * Six Selected Spill Incidents Illustrate the Various Ways Spi
          * Drinking Water Facility Operators' Opinions on Current Notif
          * Efforts Have Been Made to Supplement Existing Notification P
     * EPA's Spill Prevention and Enforcement Efforts Are Limited a
          * EPA's Prevention Program Addresses Only Oil Spills
          * EPA Is Uncertain As to Which Facilities Are Regulated by Its
          * EPA Inspections Uncovered Significant Spill Prevention Defic
          * EPA and the Coast Guard Issued a Total of 16 Penalties durin
     * Conclusions
     * Recommendations for Executive Action
     * Agency Comments and Our Evaluation
     * EPA Spill Data
     * Coast Guard Spill Data
     * Michigan DEQ Spill Data
     * Ontario SAC Spill Data
     * U.S. Spill: 2001 Hazardous Substance Spill
     * U.S. Spill: 2002 Oil Spill
     * U.S. Spill: 2004 Oil Spill
     * Canadian Spill: 2003 Hazardous Substance Spill
     * Canadian Spill: 2004 Hazardous Substance Spill
     * Canadian Spill: 2004 Oil Spill
     * GAO Contact
     * Staff Acknowledgments
     * GAO's Mission
     * Obtaining Copies of GAO Reports and Testimony
          * Order by Mail or Phone
     * To Report Fraud, Waste, and Abuse in Federal Programs
     * Congressional Relations
     * Public Affairs

Report to Congressional Requesters

United States Government Accountability Office

GAO

June 2006

CLEAN WATER

Better Information and Targeted Prevention Efforts Could Enhance Spill
Management in the

St. Clair-Detroit River Corridor

GAO-06-639

Contents

Letter 1

Results in Brief 4
Background 7
Agency Spill Data Are Not Sufficient to Accurately Determine the Actual
Number or Volume of Spills 11
Spill Notification Occurs between and among Many Different Parties and
Agreements Outlining U.S.-Canadian Notification Processes Are Not Explicit
about Time and Magnitude 17
EPA's Spill Prevention and Enforcement Efforts Are Limited and the Coast
Guard Addresses Spill Prevention as Part of Other Compliance Efforts 29
Conclusions 41
Recommendations for Executive Action 42
Agency Comments and Our Evaluation 42
Appendix I: Scope and Methodology 47
Appendix II: Spill Data for the St. Clair-Detroit River Corridor,
1994-2004 49
EPA Spill Data 49
Coast Guard Spill Data 50
Michigan DEQ Spill Data 52
Ontario SAC Spill Data 53
Appendix III: CSO Data for the St. Clair-Detroit River Corridor, 1999-2004
55
Appendix IV: NPDES Industrial Effluent Violation Data for the St.
Clair-Detroit River Corridor, 1994-2004 56
Appendix V: Spill Profiles for Six Selected Incidents 60
U.S. Spill: 2001 Hazardous Substance Spill 60
U.S. Spill: 2002 Oil Spill 61
U.S. Spill: 2004 Oil Spill 63
Canadian Spill: 2003 Hazardous Substance Spill 65
Canadian Spill: 2004 Hazardous Substance Spill 66
Canadian Spill: 2004 Oil Spill 66
Appendix VI: ORSANCO Spill Detection and Notification System 67
Appendix VII: Comments from the Department of Homeland Security 70
Appendix VIII: GAO Contact and Staff Acknowledgments 74

Tables

Table 1: Characteristics of Six Selected Spill Incidents in the Corridor
from 2001-2004 23
Table 2: Number of Coast Guard's Facility, Vessel, and Transfer
Inspections Conducted in the St. Clair-Detroit River Corridor from
1994-2004 35
Table 3: EPA Spill Data by Year and Material 50
Table 4: Coast Guard Spill Data by Year and Material 51
Table 5: Coast Guard Spill Data by Water Body and Material 51
Table 6: Michigan DEQ Spill Data by Year and Material 52
Table 7: Michigan DEQ Spill Data by Water Body 53
Table 8: Ontario SAC Spill Data by Year and Material 53
Table 9: Ontario SAC Spill Data by Water Body 54
Table 10: CSOs in the St. Clair-Detroit River Corridor from 1999-2004 by
Water Body 55
Table 11: Volume of CSOs in the St. Clair-Detroit River Corridor from
1999-2004 (in Millions of Gallons) 55
Table 12: Number of NPDES Industrial Effluent Violations by Volume per
Year 56
Table 13: Number of NPDES Industrial Effluent Violations by Discharged
Material and Year 57
Table 14: Number of NPDES Industrial Effluent Violations by Water Body and
Material for the 12 Most Frequently Violated Permits 58

Figures

Figure 1: Spill Materials in the St. Clair-Detroit River Corridor 2
Figure 2: Pollution Entering the St. Clair-Detroit River Corridor from
Sources Such as Spills, Industrial Permit Violations, and CSOs 9
Figure 3: Ontario SAC With Display Screens Showing Spill Incident
Summaries 13
Figure 4: Relative Percentages of Reported Spills, Industrial Permit
Violations, and CSOs from 1999-2004 in the St. Clair-Detroit River
Corridor That Were Greater Than 50 Gallons (or of an Unknown Volume) 15
Figure 5: Spill Notification Processes in the St. Clair-Detroit River
Corridor 18
Figure 6: Notification Processes for Six Selected Spills in the St.
Clair-Detroit River Corridor 25
Figure 7: U.S. Drinking Water Facilities on the St. Clair-Detroit River
Corridor 26
Figure 8: Existing and Proposed Water Quality Monitoring Stations in the
St. Clair-Detroit River Corridor 29
Figure 9: Number of SPCC Inspections Conducted in the St. Clair-Detroit
River Corridor from 1994-2004 33
Figure 10: EPA and Coast Guard Spill-Related Penalties Assessed from 1994
to 2004 in the St. Clair--Detroit River Corridor 38
Figure 11: 2002 Rouge River Oil Spill 61
Figure 12: 2004 Rouge River Oil Spill 63
Figure 13: ORSANCO Organics Detection System Monitoring Sites 68
Figure 14: Water Quality Monitoring Equipment Used at ORSANCO Sites 69

Abbreviations

BMP best management practices CERCLA Comprehensive Environmental Response,
Compensation, and Liability Act CSO combined sewer overflows DEQ
Department of Environmental Quality EPA Environmental Protection Agency
EPCRA Emergency Planning and Community Right-to-Know Act NPDES National
Pollution Discharge Elimination System NRC National Response Center Oil
Fund Oil Spill Liability Trust Fund OSC on-scene coordinators PEAS
Pollution Emergency Alerting System PECC Pollution Emergency
Communications Coordinator RP responsible party SAC Ontario Ministry of
Environment's Spills Action Centre SLEA Sarnia-Lambton Environmental
Association SPCC Spill Prevention, Control, and Countermeasure

This is a work of the U.S. government and is not subject to copyright
protection in the United States. It may be reproduced and distributed in
its entirety without further permission from GAO. However, because this
work may contain copyrighted images or other material, permission from the
copyright holder may be necessary if you wish to reproduce this material
separately.

United States Government Accountability Office

Washington, DC 20548

June 7, 2006 June 7, 2006

Congressional Requesters Congressional Requesters

Recent spills of large quantities of oil and hazardous substances like
vinyl chloride have raised public concerns about the safety of drinking
water for the more than 5 million people living within the St.
Clair-Detroit River corridor, which borders the United States and Canada.
The 98-mile corridor formed by the St. Clair River, Lake St. Clair, and
the Detroit River serves as a major shipping channel between the upper and
lower Great Lakes. Approximately 500 industrial facilities, including
chemical companies, oil refineries, power plants, and steel mills, are
located within the corridor. The heaviest concentration of industry on the
U.S. side is in the southern part of the corridor, around Detroit;
Canadian industry is concentrated on the northern part of the St. Clair
River, around Sarnia, Ontario. Seventeen drinking water facilities on the
U.S. side have intakes in the corridor close to these industrial centers.
The corridor is part of the Great Lakes Basin, whose waters, under the
Great Lakes Water Quality Agreement, the two countries recognized could
not be restored and enhanced without cooperation between the two
countries. Yet because of contamination from spills and other pollutant
discharges, the St. Clair and Detroit Rivers have fish consumption
restrictions, and oil sheens and other debris are commonly found on the
waters' surface or along shorelines. Recent spills of large quantities of
oil and hazardous substances like vinyl chloride have raised public
concerns about the safety of drinking water for the more than 5 million
people living within the St. Clair-Detroit River corridor, which borders
the United States and Canada. The 98-mile corridor formed by the St. Clair
River, Lake St. Clair, and the Detroit River serves as a major shipping
channel between the upper and lower Great Lakes. Approximately 500
industrial facilities, including chemical companies, oil refineries, power
plants, and steel mills, are located within the corridor. The heaviest
concentration of industry on the U.S. side is in the southern part of the
corridor, around Detroit; Canadian industry is concentrated on the
northern part of the St. Clair River, around Sarnia, Ontario. Seventeen
drinking water facilities on the U.S. side have intakes in the corridor
close to these industrial centers. The corridor is part of the Great Lakes
Basin, whose waters, under the Great Lakes Water Quality Agreement, the
two countries recognized could not be restored and enhanced without
cooperation between the two countries. Yet because of contamination from
spills and other pollutant discharges, the St. Clair and Detroit Rivers
have fish consumption restrictions, and oil sheens and other debris are
commonly found on the waters' surface or along shorelines.

Figure 1: Spill Materials in the St. Clair-Detroit River Corridor

To ensure that spills1 are promptly reported, several statutes-including
the Clean Water Act2 and the Comprehensive Environmental Response,
Compensation, and Liability Act (Superfund)3-require that the party
responsible for a discharge of oil or hazardous substance notify the
National Response Center (NRC) immediately. For oil spills, any amount
that would create a sheen on the water is a reportable quantity whereas,
for hazardous substances, Environmental Protection Agency (EPA)
regulations (prescribed under Superfund and the Clean Water Act) specify
what amounts the substances must exceed to be reportable. The NRC is the
national point of contact for spill reporting, and it distributes reported
spill information to agencies, including EPA and the Coast Guard, which
are tasked with spill response for inland and coastal spills,
respectively. In the corridor, EPA's Region 5 and the Coast Guard's
District 9 are responsible for responding to reported spills. While the
NRC is a major source of spill information for the agencies, EPA and the
Coast Guard may be contacted directly about a spill or learn of it while
conducting their regular duties. EPA and the Coast Guard's spill-related
responsibilities are outlined in the National Contingency Plan, the
federal government's guide to oil and hazardous substance spill response.
It describes how the agencies are to assess a spill, and initiate response
action if necessary. In addition to federal agencies, the Michigan
Department of Environmental Quality (Michigan DEQ) and Michigan State
Police have a role in spill-related response and notification. In Canada,
the Ontario Ministry of Environment has a prominent role in spill-related
activities in the St. Clair-Detroit River corridor, including notification
and enforcement. The Ontario Ministry of Environment's Spills Action
Centre (Ontario SAC) obtains and reports spill information to responders
and other stakeholders. Environment Canada, the federal-level regulatory
agency, has a lesser role in spill-related activities in the corridor.

1For purposes of this report, "spill" refers to an unanticipated or
accidental discharge of pollutants into the waters of the corridor, but
excludes National Pollutant Discharge Elimination System (NPDES) permit
violations and combined sewer overflows. However, we treat permit
violations and combined sewer overflows reported to the NRC as spills
within this report.

2Federal Water Pollution Control Act ("Clean Water Act"), 33 U.S.C. S:S:
1251-1387.

3Comprehensive Environmental Response, Compensation, and Liability Act
("Superfund"), 42 U.S.C. S:S: 9601-9675.

To decrease the likelihood of spills, both EPA and the Coast Guard
administer spill prevention programs. EPA's program requires that
non-transportation related facilities (e.g., power plants) with a
specified storage capacity prepare spill prevention plans and meet certain
operational standards. These facilities must also provide containment
structures or explain in their plans why the structures are not feasible.
The Coast Guard's prevention program requires vessels with the capacity to
carry certain amounts of oil and hazardous substances-and certain
facilities that transfer oil and hazardous substances-to have written
transfer procedures or operations manuals, meet minimum equipment
requirements, and follow prescribed procedures when transferring oil on
navigable waterways. EPA and the Coast Guard have the authority to fine
parties that are responsible for spills, or that do not comply with spill
prevention program requirements.

In this context, you asked us to examine (1) how many oil and hazardous
substance spills of more than 50 gallons (or of an unknown volume) were
reported in the St. Clair-Detroit River corridor from 1994 to 2004, and
how accurately reported spills reflect the extent of actual spills; (2)
what processes are used to notify parties of spills, and whether they
contain explicit requirements for reporting times and spill magnitude; and
(3) the extent of EPA and the Coast Guard's spill prevention efforts and
enforcement activities in the St. Clair-Detroit River corridor from 1994
to 2004.

To obtain overall information relating to spills, we gathered information
from officials within various offices in EPA; the Coast Guard; the state
of Michigan; and Ontario, Canada. To determine the number of spills over
50 gallons (or of an unknown volume) reported in the St. Clair River, Lake
St. Clair, and the Detroit River and its tributary, the Rouge River,
between 1994 and 2004 we obtained spill data maintained by NRC, EPA, the
Coast Guard, Michigan DEQ, and the Ontario SAC. We also obtained EPA and
Michigan DEQ data sets related to other pollutant discharges, such as
combined sewer overflows (CSO) and National Pollution Discharge
Elimination System (NPDES) industrial permit violations to obtain more
complete information on pollutants discharged into the water bodies of the
St. Clair-Detroit River corridor. To assess what processes are used to
notify parties of spills and whether they contain explicit requirements
for reporting times and spill magnitudes, we reviewed applicable laws and
spill notification agreements and obtained information on implementation
of these agreements. We obtained and analyzed documentation on six spills
to better understand how the notification process was conducted under
varying circumstances, and to show a range of spill materials and source
locations. We also obtained information from all 17 of the drinking water
facility operators on the U.S. side of the corridor to obtain their
perspectives on the timeliness of spill notification. To identify spill
prevention efforts and enforcement activities, we obtained and analyzed
the relevant legislation, agency spill prevention and enforcement
policies, as well as agency data on spill-related inspections and
enforcement actions in the corridor since 1994. We performed our work
between September 2005 and June 2006 in accordance with generally accepted
government auditing standards. A more detailed description of our
methodology is provided in appendix I.

                                Results in Brief

The NRC received 991 spill reports from the U.S. side of the corridor and
the Ontario SAC received an additional 157 spill reports from the Canadian
side of the corridor from 1994 to 2004, but these reports do not
accurately capture the actual number or volume of spills. U.S. and
Canadian officials believe that many spills go unreported because, among
other reasons, responsible parties may not understand or comply with the
reporting requirements. In addition, the NRC database often includes
multiple reports for the same spill. The response agencies-EPA and the
Coast Guard-are required to assess each reported spill and therefore
should have reliable spill information based on their investigations, but
this is not the case. EPA Region 5 does not remove all duplicate spill
reports, or update its data after investigating spills. Furthermore,
investigators do not consistently document their activities, so it is
unclear which spills have or have not been investigated. In contrast,
Coast Guard officials in District 9 document their investigations and use
the information to update their spill data, but they do not update spill
volume estimates after investigations because of automated system
limitations that do not allow them to input revised estimates. As a
result, complete information on the extent of spills is not available to
determine if additional management or oversight is needed to address the
problem. Finally, we also found that, according to agency data sets, other
events-CSOs and industrial permit violations-occurred more frequently than
spills in the corridor.

There are multiple parties involved in spill notification in the corridor
and agreements outlining U.S.-Canadian notification processes are not
explicit about reporting times or the magnitude of spills that warrant
notification. Generally, spill notification involves the following: (1)
spill occurrence and reporting by a responsible party or observer to a
designated reporting center or a response agency; (2) spill notification
from response agencies to one another if determined to be necessary; and
(3) spill notification by response agencies to drinking water facilities
and other stakeholders. All spill notification is dependent upon reporting
by parties responsible for the spill or others who provide initial
notification of the spill. Spill notification processes between the United
States and Canada are outlined in two agreements between the countries.
One process is used by the coast guards of each country, who have agreed
to notify one another of spills primarily when a joint response may be
needed. Another process is used at the state and provincial level, where
officials from the Michigan State Police and Ontario Ministry of
Environment have an agreement to notify one another of spills. However,
these two agreements are not explicit about what constitutes timely
reporting or what magnitude of spill triggers notification. According to
Michigan State Police officials, the Michigan-Ontario notification
agreement was meant to, among other things, expedite the dissemination of
information to the public, because responsibility for notifying the
public-including drinking water facilities-generally resides with state
and local authorities. We reviewed six selected spill incidents to gain
insight into the dynamics of spill notification processes from initial
reporting to drinking water facility notification. The drinking water
facility operators we contacted on the U.S. side of the corridor had
differing perspectives on current notification processes, and the majority
expressed concern that their facilities could be contaminated by spills if
notification was not timely. Finally, efforts have been made to develop
informal notification processes between individual industries or trade
associations and drinking water facilities.

EPA's spill prevention efforts are limited to potential oil spills,
whereas the scope of the Coast Guard's efforts are broader and included as
part of other compliance activities. EPA's Spill Prevention, Control, and
Countermeasure (SPCC) program addresses only oil spills, not spills of
other hazardous substances. Agency officials told us that their spill
prevention regulations were written when oil spills were a bigger problem,
and they believed that the NPDES program already addressed chronic
pollution discharges in waterways. Unlike EPA's NPDES program, which
requires facilities to obtain a permit to discharge pollutants, the SPCC
program has no mechanism to identify the facilities that it regulates.
This creates a challenge in determining which facilities are required to
create spill prevention plans. To date, EPA has not identified the
universe of facilities to be inspected under its SPCC program. EPA Region
5 officials have identified 59 facilities in the corridor that are
regulated under the SPCC program as a result of referrals from Michigan
DEQ or through special inspection initiatives. Inspections for SPCC
program compliance vary in number per year and are largely contingent upon
other inspection initiatives because of limited numbers of inspection
staff, according to agency officials. The inspections that were conducted
disclosed significant and numerous spill prevention deficiencies, such as
failure to prepare prevention plans. On the other hand, the Coast Guard
inspected a greater number of the facilities and vessels under their
jurisdiction than EPA-mainly because the inspections were not only focused
on spill prevention, but included other areas related to safety and
security. Officials from the Coast Guard reported that in nearly all
cases, their inspectors found that facilities and vessels were in
compliance; those that were not in compliance had only minor issues, such
as incidental omissions in operations manuals. In response to violations
for spills or program noncompliance, EPA and the Coast Guard issued a
total of 16 penalties. From 1994 through 2004, EPA Region 5 fined four
parties an average of $39,000 each; three were for violations of the spill
prevention program and one was for a spill. In assessing reported spills,
EPA relies on Michigan DEQ to respond locally and does not respond on-site
to the majority of spills reported to them. Therefore, it does not collect
evidence from many reported spills that could be used to take enforcement
actions against responsible parties. In contrast, officials from the Coast
Guard told us they investigate all spills reported to them, both on-site
and indirectly (e.g., through inquiries). Based on these spill
investigations, the Coast Guard's District 9 fined 12 parties an average
of about $2,100 each in the time frame we reviewed. Neither agency fined
parties for failure to report spills.

To improve spill management, we are making recommendations to EPA and the
Coast Guard aimed at providing better documentation of spill investigation
results, and identification of possible enhancements to the notification
process. We are also recommending that EPA develop an inventory of
facilities that are subject to its spill prevention program and develop
goals for the frequency and extent of its inspections for such facilities.

In commenting on a draft of this report, the Department of Homeland
Security (DHS) agreed overall with our findings and conclusions, and EPA
provided technical comments only. DHS did not address our recommendations.
In commenting on our recommendation that it gather information on
SPCC-regulated facilities, EPA stated that there is no authority in the
Clean Water Act or the prevention regulations for facilities to provide
this information to EPA. Furthermore, it stated that under the Paperwork
Reduction Act the agency would need to seek approval from the Office of
Management and Budget to collect such information. In this regard, we note
that EPA has previously identified SPCC-regulated facilities in the
corridor. However, if EPA determines that formal rulemaking is necessary
for it to gather information on which facilities are covered under its
spill prevention program, then we believe it should consider undertaking
such a rulemaking. EPA also commented on the feasibility of updating spill
information maintained by the NRC. While we acknowledge that EPA does not
modify spill data maintained by the NRC, our recommendation was that EPA
update its own spill data and explore the feasibility of updating spill
information maintained by NRC by informing the NRC of duplicate spill
reports.

                                   Background

The St. Clair and Detroit Rivers, and Lake St. Clair, provide multiple
benefits to residents of Michigan and Ontario, Canada, who use the water
bodies as their primary source of drinking water as well as for recreation
such as boating and fishing. Sensitive ecological areas located along the
corridor include Humbug Marsh, the last Great Lakes coastal marsh on the
Michigan mainland of the Detroit River. It contains the greatest diversity
of fish species found in the Detroit River and it is part of the migration
route for 117 fish and 92 bird species. The Detroit River itself was
designated an American Heritage River in 1998 for these ecological
resources. Despite these and other benefits, the St. Clair and Detroit
Rivers are considered "Areas of Concern" by the U.S. and Canadian
governments under the Great Lakes Water Quality Agreement as a result of
beneficial-use impairments, such as restrictions on fish consumption.

Pollutant discharges to the waters of the corridor include CSOs-caused by
heavy rains that force wastewater treatment plants to bypass their
overburdened systems and discharge raw or partially treated waste directly
into the water bodies. Michigan law requires that wastewater treatment
facilities report their combined and sanitary sewer overflows to the
Michigan DEQ within 24 hours. Discharges from industrial facilities with
NPDES permits account for additional pollutants that enter the waters of
the corridor. Industries with NPDES permits are required to report on the
quality of all discharges and to detail any pollutants discharged that
exceed their permit limits to EPA in monitoring reports at intervals
specified in their permits, commonly monthly. As a result, NPDES-permitted
industries regularly monitor their discharges. In addition to these
requirements, federal law requires that parties that discharge oil or a
hazardous substance beyond specified quantities into waters of the
corridor report these incidents to the NRC. Spills and other pollutant
discharges might also be reported to the NRC by members of the public that
observe pollutant materials in waterways. When spills, industrial permit
violations, and sewer overflows contain oil, they are visible-and more
likely to be reported by observers. In contrast, releases of chemicals
into the water are oftentimes not visible, unless they can be detected by
their effects, such as fish kills. Figure 2 illustrates these sources of
pollution.

Figure 2: Pollution Entering the St. Clair-Detroit River Corridor from
Sources Such as Spills, Industrial Permit Violations, and CSOs

While EPA has federal regulatory responsibility for NPDES-related
discharges and CSOs, EPA and the Coast Guard share responsibility for
spill prevention and response on the U.S. side of the corridor. The
National Contingency Plan and the Southeast Michigan Area Contingency Plan
describe a geographic division of responsibility between these agencies,
but due to EPA's expertise, the Coast Guard may refer chemical spills to
EPA even if the spills are in locations otherwise assigned to the Coast
Guard.4 When spills originate on land but impact the navigable waters,
both agencies might be involved in response. Within EPA, Chemical
Preparedness officials enforce regulations that address chemical release
reporting requirements while Oil Program officials coordinate spill
response and oil spill prevention inspections. When spills involve
industrial permit violations or sewage releases, EPA's NPDES program
officials are also involved-but because EPA approved Michigan's NPDES
program, Michigan officials are more directly involved in these cases. As
agencies respond to spills, they work with responsible parties to ensure
that they fund the cost of cleanup activities. If EPA and the Coast
Guard's spill responders do not identify the responsible party, however,
they may obtain funds from the Oil Spill Liability Trust Fund (Oil Fund)
or Superfund to finance their response efforts, including the cleanup.

According to officials from the Coast Guard, notification of potentially
affected parties is oftentimes a component of the agencies' spill response
efforts. In addition to federal agencies, the Michigan DEQ and State
Police also provide spill notification. On the Canadian side of the
corridor, the Ontario SAC consolidates spill reports routed to their
center, as well as to other agencies. For example, the Ontario Ministry of
the Environment has an agreement with Environment Canada under which they
receive all spill reports for the federal agency. There are many potential
pathways for spill notification in the corridor. The overall process can
be divided into spill occurrence and reporting by a responsible party or
observer to a designated reporting center; spill reporting from designated
spill reporting centers to response agencies; and spill notification from
response agencies to stakeholders, including drinking water facilities.
Sometimes parts of the process are collapsed; for example, spill reporting
centers may notify other stakeholders as well as response agencies.
Alternatively, the process can be lengthened if multiple agencies are
responsible for notifying other stakeholders in sequence.

4In accordance with Section 10(g) of Executive Order 12580 of January 23,
1987, Coast Guard officials may use an instrument of re-delegation to
shift leadership of response efforts to EPA.

 Agency Spill Data Are Not Sufficient to Accurately Determine the Actual Number
                              or Volume of Spills

Agency spill data are not sufficient, for multiple reasons, to accurately
determine the actual number or volume of spills in the St. Clair-Detroit
River corridor. Many spills go unreported because responsible parties may
not understand or comply with reporting requirements. On the other hand,
there are oftentimes multiple NRC reports for the same spill, since
several observers may report them. EPA Region 5 does not remove all
duplicate spill reports from their database, or update its data after
investigating spills. In contrast, Coast Guard officials in District 9
document their investigations and use the information to update their
spill data, but they do not update spill volume estimates because of
automated system limitations. Other events, including CSOs and industrial
permit violations, are reported more frequently in the corridor.

Agency Officials Believe Responsible Parties May Not Report All Spills

NRC, EPA, Coast Guard, and Canadian officials believe that many spills are
never reported, and therefore that spill data do not represent the true
number of spills. Though responsible parties are required by law to
immediately report spills in amounts beyond certain minimum quantities,
agency officials believe they may not do so for a variety of reasons.5
U.S. and Canadian officials suggested that responsible parties may not be
aware of spills, may not understand the reporting requirements, or that
they may not want to receive "bad press" or be forced to pay the costs of
the cleanup. Reporting by responsible parties and others is critical
because only one water quality monitoring station capable of detecting
spills exists in the corridor. The Sarnia-Lambton Environmental
Association (SLEA), a Canadian industry consortium, maintains a monitoring
station south of the highly industrialized Sarnia area. Though SLEA
monitors for a suite of chemicals, it does not detect all types of
discharges-and while it shares spill data with the Ontario Ministry of
Environment, its purpose is not to collect spill-related information for
regulatory agencies; rather, it collects the information as a service to
SLEA members, as well as agencies and communities.

5Responsible parties are required to report spills by the Clean Water Act
and Comprehensive Environmental Response, Compensation, and Liability Act
(CERCLA); Ontario's Environmental Protection Act also requires that
parties report spills.

When spills are reported, in many cases the responsible party is unknown.
In many of these instances, a member of the public or party other than the
responsible party provides information to the NRC. EPA and the Coast
Guard's spill data indicate that 67 percent and 29 percent, respectively,
of reported spills in the corridor were released from an unknown source in
the time period we reviewed. Ontario SAC data indicate that 10 percent of
Canadian spills were from an unknown source.

NRC Data Contains Multiple Reports of the Same Spills

Another reason spill data do not accurately represent the number of actual
spills is that NRC spill data record some spill events multiple times. The
NRC received 991 reports of spills in the corridor from 1994 to 2004, but
these may include multiple reports of the same spills.6 NRC officials are
responsible for maintaining a call center for obtaining spill information,
and relaying the information to the appropriate agencies that are tasked
with response. They are not required to assess whether multiple reports
pertain to the same spill, as this would require investigation. NRC
officials told us that, as a result, many duplicate spill reports exist.
Coast Guard officials from District 9 told us that they could, after
investigating spill incidents, identify duplicate spill reports provided
by NRC, link these duplicate reports to single spill incidents, and
provide that information to the NRC so that they can update their records.

Duplicate reporting has been addressed by Ontario's SAC, which obtained
157 reports of spills on the Canadian side of the corridor during the same
time period. Unlike the NRC, the Ontario SAC determines whether each spill
report is unique when it records its information. The Ontario SAC is
staffed by Ontario Ministry of Environment officials who are responsible
both for obtaining preliminary spill information for the province, as well
as for determining which spill reports relate to the same incident. The
Ontario SAC has a rolling summary of spill incidents on a display screen
and on staff computers, which allows them to identify multiple reports
that relate to a common incident. (See fig. 3.) The Ontario SAC's
Emergency Management Coordinator told us that when these safeguards fail
to eliminate a duplicative spill report, subsequent corrections are made.

6The NRC spill data include spills greater than 50 gallons (or of an
unknown volume).

Figure 3: Ontario SAC With Display Screens Showing Spill Incident
Summaries

To develop a process similar to Ontario's for the U.S. side of the
corridor, Michigan State Police officials told us that between 1986 and
1988 state officials explored the option of creating a spills center. At
the time, they estimated that it would have cost $2 million to operate and
it would have required 10 staff, including a chemical specialist and three
shifts of phone operators. This was viewed as prohibitively expensive by
Michigan officials, and as an alternative, the State Police and the
Michigan DEQ's Pollution Emergency Alerting System (PEAS) began operating
as a spill notification system. The PEAS system is used for reporting
spills to the Michigan DEQ during non-business hours, including holidays,
weekends, and evenings. Spill data from PEAS, however, are similar to NRC
data in that they include multiple entries for single spills because each
call is logged, rather than each unique spill event recorded.

EPA Does Not Eliminate All Duplicate Spill Reports or Update Spill Information
after Investigations

Unlike the NRC, response agencies such as EPA are required to assess each
reported spill and therefore should have reliable spill information, but
this is not the case. EPA Region 5 does not eliminate all duplicate spill
reports because they do not respond on-site to the majority of spills for
which they receive reports. EPA Region 5 officials told us that they rely
on Michigan DEQ to respond to the majority of spills since they are in
closer proximity. Region 5 officials respond to spills upon receiving a
request for assistance from Michigan DEQ officials, and when spills are
over 1,000 gallons, EPA officials respond to provide assistance even if
they are not requested to do so. They told us that they investigate very
few spills on-site-perhaps roughly one percent of spills-due to limited
staff resources. Instead, EPA Region 5 officials follow up with state
spill responders by phone to obtain more detail on spills. Though their
operating protocols state that responders are to complete pollution
reports and update spill data after investigation, EPA Region 5 officials
told us that responders have not done so typically because they fail to
make it a priority. For this reason, EPA officials were unable to tell us
which spills in the corridor in our time frame were investigated by their
agency. They told us that EPA imports spill data from the NRC and does not
make modifications to the data; therefore, EPA's spill data set is of
limited use. EPA Region 5 officials providing spill response in the
corridor began using a new Web-based spill data system, Web Emergency
Operations Center (Web EOC), in the fall of 2004. EPA officials are
hopeful that spill responders will update spill information in the system
following their investigations; however, they said that it is too soon to
tell.

The Coast Guard Updates Most Spill Data but Does Not Update Spill Volumes

Like EPA, Coast Guard officials from District 9 told us that they assess
and investigate each spill, whether they go on-site or use phone calls and
other means to obtain information; however, Coast Guard officials update
spill information following these investigations. While the Coast Guard's
spill data sets included information on spill materials, the cause of
spills, and how each spill was resolved, the formatting of the data sets
makes it difficult to access accurate information on spill volumes. For
example, a spill listed in the Coast Guard's data set as being a
2,000-gallon spill is also reported in the Coast Guard's annual report as
being over 8 million gallons. Similarly, the Coast Guard's spill data set
contains a reference to a 75-gallon oil spill, but summaries written by
the Coast Guard's District 9 responders to the spill state that over
66,000 gallons of oil were recovered. When asked about the discrepancies
in these cases and others, Coast Guard officials from District 9 told us
that they are unable to update the field in their database that contains
preliminary volume estimates. Instead, they update volume information in
narrative fields. As a result, it is difficult to assess the severity of
any given spill in the Coast Guard's data sets.

Number of Reported CSOs and Industrial Permit Violations Exceed the Number of
Reported Spills in the Corridor

The number of reported spills is exceeded by other types of events, such
as CSOs and industrial permit violations that are reported more frequently
in the corridor. EPA's data on U.S. industrial permit violations indicate
that approximately 2,200 were reported in the corridor during the 11-year
time period we reviewed; over 1,800 were greater than 50 gallons (or of an
unknown volume). Michigan DEQ has tracked CSOs on the U.S. side of the
corridor since 1999. Their data indicate that roughly 1,400 CSOs were
reported in the corridor from 1999 to 2004. These data might be subject to
the same limitations as the spill data because industrial permit
violations and CSOs are self-reported and facilities may not report all of
these events. However, spills may be particularly subject to
underreporting because they are not part of a structured program-as CSOs
and industrial permit violations are. Figure 4 illustrates the relative
percentages of spills, industrial permit violations, and CSOs of greater
than 50 gallons (or of an unknown volume) that were reported in the
corridor in the 6-year period between 1999 and 2004, the time period for
which CSO data were available.

Figure 4: Relative Percentages of Reported Spills, Industrial Permit
Violations, and CSOs from 1999-2004 in the St. Clair-Detroit River
Corridor That Were Greater Than 50 Gallons (or of an Unknown Volume)

Notes:

(1) Percentages do not total 100 due to rounding.

(2) This chart depicts the frequency, rather than the toxicity, of
pollutant events. The data used to determine relative frequencies are
limited by their reliance upon self-reporting by facilities.

Typically, CSOs in the corridor contain biological waste, commercial and
industrial waste, and storm water runoff from streets and other surfaces.
In the Detroit area, however, CSOs are more likely to contain industrial
waste in concentrations that have the potential to negatively impact water
quality to a greater extent. In addition to sewage from 3 million area
customers and 78 municipalities that send their waste to the Detroit
plant, the wastewater treatment facility treats industrial waste from over
250 major industries. The facility has approximately 80 outfalls and is
one of the largest wastewater treatment plants in the world.7 While the
facility has an industrial pretreatment program that requires that
industries' waste meets certain limits before treatment, these limits may
be relatively lenient, according to EPA officials, resulting in high
volumes of waste flowing into the facility. For example, EPA officials
told us that the facility has lenient oil and grease pretreatment limits.
In the event of a CSO, the pretreated material that bypasses the Detroit
wastewater treatment facility and is discharged into the Detroit and Rouge
Rivers may contain industrial waste, including oil, grease, and other
materials. The Detroit facility has historically had difficulties
complying with permit requirements. To address these deficiencies, EPA
filed suit against the Detroit facility in the 1970s and the resulting
consent decree has, according to EPA officials, provided a basis for many
required changes to improve their facility.8 However, a lawsuit filed by
EPA in the 1980s which related primarily to the facility's industrial
pretreatment program was dismissed in federal court.9

7An outfall is any pipe or conduit used to carry either raw sewage or
treated effluent to a final point of discharge into a body of water.

8United States v. City of Detroit, 476 F. Supp. 521 (E.D. Mich. 1979).

9United States v. City of Detroit, 940 F. Supp. 1097 (E.D. Mich. 1996)
(granting Detroit's motion for summary judgment).

     Spill Notification Occurs between and among Many Different Parties and
Agreements Outlining U.S.-Canadian Notification Processes Are Not Explicit about
                               Time and Magnitude

Spill notification may involve the following: (1) spill occurrence and
reporting by a responsible party or observer to a designated reporting
center or a response agency; (2) spill notification from response agencies
to one another; and (3) spill notification by response agencies to
drinking water facilities and other stakeholders. Spill notification
between the United States and Canada is outlined in two agreements.10 The
coast guards of each country and officials from the Michigan State Police
and Ontario SAC have agreed to notify one another of spills; however,
these two agreements are not explicit about which spills warrant
notification or how quickly notification should occur. We reviewed six
selected spill incidents to gain insight into the spill notification
process from initial reporting to drinking water facility notification.
Drinking water facility operators on the U.S. side of the corridor had
differing perspectives on current notification processes, but the majority
expressed concern that their facilities could be contaminated by spills
due to untimely notification. Finally, efforts have been made to develop
informal notification processes between individual industries or trade
associations and drinking water facilities.

Spill Notification Is Multi-Faceted and May Follow Many Different Pathways

There are several potential pathways through which spill notification may
occur in the corridor. The overall process can be divided into spill
occurrence and reporting by a responsible party or observer to a
designated reporting center or response agency; spill reporting from
designated spill reporting centers to response agencies; spill
notification from response agencies to other response agencies; and
notification to stakeholders, including drinking water facilities.
Sometimes parts of the process are collapsed; for example, spill reporting
centers may notify other stakeholders as well as response agencies.
Alternatively, the process may be lengthened if multiple agencies are
responsible for notifying other stakeholders in sequence (see fig. 5).

10In addition to these spill notification agreements related to marine
spills, the Canada-U.S. Joint Inland Pollution Contingency Plan provides
for notification for inland spills. Environment Canada's National
Environmental Emergencies Centre notifies the NRC and vice versa. Canadian
officials told us the plan, as implemented, addresses mainly air releases.

Figure 5: Spill Notification Processes in the St. Clair-Detroit River
Corridor

Two Agreements Outline Notification between the United States and Canada

The Canada-United States Joint Marine Pollution Contingency Plan states
that on-scene coordinators (OSC) from the U.S. and Canadian Coast Guards
may notify each other of spills when there is a substantial threat of the
spreading of pollutants across shared boundaries, including the St.
Clair-Detroit River corridor and other waters of the Great Lakes. The plan
arises from the Great Lakes Water Quality Agreement between Canada and the
United States that calls for development of a joint contingency plan for
use in the event or threat of a spill involving oil or a hazardous
substance. The notification called for in the plan is conducted by phone
between the two coast guards. They also provide warning messages to each
other when they are uncertain as to whether a spill will impact the
other's waters; when a joint response is needed to address a spill, they
call or communicate via fax. Officials from the Coast Guard told us the
plan has only been utilized for joint response twice since 1994. While
spill-related warnings have not been systematically tracked between the
U.S. and Canadian Coast Guards, officials from the U.S. Coast Guard told
us they are starting to track the warning messages to and from Canada.
Though U.S. Coast Guard officials may notify Canadian Coast Guard
officials of spills, there is no guidance or directive for either party to
notify local stakeholders, such as drinking water facilities; however,
they told us that they sometimes do so as a courtesy.

Though the U.S. and Canadian Coast Guards have had a spill notification
process in place since 1978, Michigan and Ontario officials believed that
another notification process was necessary at the state and provincial
level to expedite notification of stakeholders such as drinking water
facilities. To address this need, the State of Michigan and Province of
Ontario agreed in 1988 to contact one another by phone if an unanticipated
or accidental discharge of pollutants occurred and the discharge was
likely to adversely affect the adjoining jurisdiction or drinking water
supply. Michigan State Police were designated as the authority responsible
for this task by the state governor because they have the capability to
receive information on a 24-hour basis, 7 days a week. According to
Michigan State Police officials, this notification process was intended to
provide immediate spill-related information to state authorities, who in
turn could provide that information to stakeholders such as drinking water
facilities. These officials told us that they believe that duplication in
notification efforts at the federal and state levels is beneficial,
because stakeholders at all levels are more likely to obtain information
if multiple processes are involved, since any one system might fail.

The responsibility for communicating spill information to the public
generally resides with state and local authorities, who are presumed to be
the first agencies on the scene. This responsibility was established in
the Emergency Planning and Community Right-to-Know Act (EPCRA) of 1986,11
which requires states to establish an emergency planning and notification
system. This system includes local emergency planning commissions, which
are charged with creating procedures for receiving and responding to
public requests for information. However, there is no proactive
notification requirement in the act for the local planning commissions.

Neither Agreement is Explicit about When or How Quickly Notification Should
Occur

Neither the Joint Marine Pollution Contingency Plan nor the
Ontario-Michigan Joint Notification Plan contains explicit requirements
for what types of spills warrant notification or how quickly notification
must be given. For example, Ontario Ministry of Environment officials told
us that they classify some sewer overflows as spills. These include sewage
bypasses caused by equipment failure, power outages, and maintenance
shutdowns. Michigan officials, on the other hand, do not consider these
events to be spills because they are regulated separately. U.S. Coast
Guard officials said they do not regularly provide information about sewer
overflows to Canadian officials, since they are not required to do so,
these events occur too frequently, and it would not be feasible to relay
information on each occurrence in the corridor. Even when Ontario and
Michigan officials agree on what type of event is considered a spill, they
told us that they do not have a common understanding of what magnitude of
spill requires notification. According to Michigan officials, the
agreement does not specify spill volumes that trigger notification,
because the agreement's authors were more concerned with spill-material
toxicity. Michigan and Ontario officials told us that they have tried to
better define when notification is required, but they are frustrated
because they have not yet reached consensus on the issue. For example,
Michigan officials independently explored the idea of notifying Ontario
officials of spills only when spills exceeded 1,000 gallons. Ontario
officials, upon learning of this limit, did not agree. They thought this
figure was too high and also indicated that volume alone is not an
adequate measure of potential impact. In their opinion, other factors such
as toxicity and concentration also need to be considered. Since two large
chemical spills occurred at Sarnia industrial facilities in 2003 and 2004,
Ontario officials told us they have notified Michigan officials of spills
of various sizes but have not always been informed of large U.S. spills by
Michigan authorities. Ontario officials provided some examples of when the
Province learned of spills originating in Michigan and impacting Ontario
through calls to the SAC from fisherman and other local stakeholders.
Michigan State Police officials told us they are uncertain as to whether
they are notified of all Canadian spills. These officials have not tracked
spill notification to and from Ontario; however, they told us they intend
to start doing so.

11Pub. L. No. 99-499, tit. III, 100 Stat. 1728 (1986).

Though the Ontario-Michigan spill notification agreement specifies that
notification is to be immediate for those spills likely to adversely
affect the adjoining jurisdiction, officials on both sides told us that
they are not always notified in a timely manner. Michigan DEQ officials
told us that the greatest lag in the notification process is the time
between when a spill occurs and when it is reported by a responsible party
to agency officials. Ontario officials told us that they are not always
able to notify immediately because some assessment is often required to
determine if there is any likelihood of an impact on the U.S. side.
Ontario officials also told us that the number of parties or steps
involved in the Michigan notification process is greater than those
involved in their process, and this could contribute to delays in
Michigan's spill notification. A local official from a county bordering
Lake St. Clair also told us that the process employed by Michigan State
Police and Michigan DEQ officials to notify stakeholders has too many
steps, and drinking water facilities are too far down on their list for
timely notification. Two local officials told us that Michigan's spill
notification process should include electronic communication, rather than
relying exclusively on a phone tree, since this provides too many
opportunities for communication to be disrupted.

Six Selected Spill Incidents Illustrate the Various Ways Spill Notification Can
Occur

Spill notification varies from spill to spill, depending on the unique
circumstances of the incident. We selected six spill cases to illustrate
the various ways that spill notification can occur. These six cases were
chosen to maximize variability among several factors including country of
origin, spill material, and whether the responsible party was known. (See
table 1.)

Table 1: Characteristics of Six Selected Spill Incidents in the Corridor
from 2001-2004

         Country                                                                       
Month/   of                          Volume         Responsible             
year     origin  Location Material   estimate       party       Cause       Impacts
May 2001 United  Detroit  Ethylene   8.4 million    Known       Pipe from   Wildlife   
         States  River    Glycol and gallons (CGa)              containment injury and 
                          Propylene  25 million                 pond to     death      
                          Glycol     gallons storm              sewer       
                                     water/ethylene             system      
                                     glycol mixture             became      
                                     (RPb)                      blocked     
April    United  Rouge    Oil/sewage 321,000        Unknown     CSO-related Wildlife   
2002     States  and      mixture    gallons of oil                         injury and 
                 Detroit             (EPA)                                  death      
                 Rivers                                                                
                                     > 66,000                               Vessel     
                                     gallons of oil                         transits   
                                     recovered (CG)                         delayed,   
                                                                            cancelled, 
                                                                            and        
                                                                            diverted   
                                                                                       
                                                                            U.S. and   
                                                                            Canadian   
                                                                            shorelines 
                                                                            oiled      
August   United  Rouge    Oil/sewage 5,000 gallons  Unknownc    CSO-related None       
2004     States  and      mixture    (CG)                                   documented 
                 Detroit                                                    
                 Rivers                                                     
August   Canada  St.      Vinyl      34 and 5       Known       Cracked     None       
2003             Clair    chloride   gallon spills              tube in     documented 
                 River    monomer    (RP)                       heat        
                                                                exchanger   
February Canada  St.      Methyl     > 39,000       Known       Leaking     Drinking   
2004             Clair    ethyl      gallons (RP)               heat        water      
                 River    ketone and                            exchanger   facility   
                          methyl                                            intake     
                          isobutyl                                          closures   
                          ketone                                            
May 2004 Canada  St.      Oily water Unknown        Known       Rain caused None       
                 Clair                                          oil         documented 
                 River                                          separator   
                                                                overflows   

Source: EPA, U.S. Coast Guard, Ontario SAC, and Michigan DEQ.

aCG is an abbreviation for the Coast Guard.

bRP is an abbreviation for Responsible party.

CIn this case, the City of Detroit was held accountable for the spill and
another that occurred later that month, because both originated in their
sewer system.

In three of the six cases we reviewed, the public, rather than the
responsible party, was the first source of spill information to response
agencies. In one of these cases, the responsible party later provided the
approximate time that the incident occurred and therefore we could
calculate the time between spill occurrence and reporting, which was about
24 hours. Notification of agency officials and then drinking water
facilities occurred most quickly when the responsible party reported the
spill within 2.5 hours of its occurrence. In February and May of 2004, a
spill of methyl ethyl ketone and then an oily water spill occurred in
Ontario; these entered the St. Clair River. For these spills, Ontario
officials notified Michigan officials within 1 to 2 hours of the spill
being reported. Michigan drinking water facilities were then informed of
the spill by Michigan officials within the next 1 to 2.5 hours. For these
incidents, notification took less than 5 hours, from spill occurrence to
notification of drinking water facilities. When responsible parties did
not promptly report the spill, the notification process took 2 days or
more. For two chemical spills that we reviewed, including an ethylene and
propylene glycol spill in Michigan and a vinyl chloride monomer spill in
Ontario, the responsible party failed to notify regulatory officials until
several days after the spill occurred. The Canadian spill was not detected
by the responsible party, because their monitoring equipment was not
running as a result of a power outage. The U.S. spill was not detected
until a member of the public observed fish dying and reported it to
Michigan DEQ officials; the responsible party failed to notify state
officials of the spill. In addition, our review of six selected cases
illustrated that in five cases, agencies notified one another per the
notification agreements. In the case in which they did not, Michigan
officials determined that there was no potential impact to Canadian
waters. Finally, for the six spills we reviewed, drinking water facilities
were not notified in three instances. In these cases, agency officials
determined that it was unnecessary to notify the facilities because, in
their view, the facilities would not be affected or the information was
deemed too late to be useful. Figure 6 shows the notification milestones
for the six spills we profiled.

Figure 6: Notification Processes for Six Selected Spills in the St.
Clair-Detroit River Corridor

aIn these cases, the responsible party provided the approximate time the
spill occurred.

bFor the April 2002 and August 2004 spills, the responsible party was not
identified and the time the spills occurred is unknown.

Drinking Water Facility Operators' Opinions on Current Notification Processes
Varied but Many Expressed Concern That Potential for Plant Contamination Exists

Drinking water facility opinions varied-by location along the
corridor-about the timeliness of spill notification. While nearly all
drinking water facility operators with facilities along the St. Clair
River and northern half of Lake St. Clair told us that spill notification
was not timely, almost all facility operators with facilities along the
lower half of Lake St. Clair and the Detroit River told us that
notification was timely. These facility operators indicated that proximity
to spill locations makes a difference in their definition of notification
timeliness because they might have more or less time to prepare for spill
material to pass their intakes. Figure 7 illustrates the location of U.S.
drinking water facilities in the corridor. Despite the difference of
opinion on notification timeliness, the majority of the 17 drinking water
facility operators all along the corridor told us they would like to be
notified of a spill immediately, or within 1 hour or less of its
occurrence. In the six spills we profiled, notification never occurred in
this time frame.

Figure 7: U.S. Drinking Water Facilities on the St. Clair-Detroit River
Corridor

Furthermore, many Michigan drinking water facility operators along the
corridor expressed concern that their facilities could be contaminated by
spills. Some cited factors that could increase the likelihood of facility
contamination, such as vessel traffic along the corridor or the number of
industries located along the corridor. They told us that spill
notification plays a key role in whether their facilities might be
contaminated. Some told us that spill notification is the most important
factor in their ability to protect the drinking water. Two facility
operators also indicated that their customers have expressed concerns
about the safety of their drinking water. Generally, facility operators
located along the St. Clair River and the top of Lake St. Clair seemed to
express greater concern than facility operators located along the southern
part of Lake St. Clair and the Detroit River. For example, a facility
operator in the northern part of the corridor told us that he believes
drinking water facility contamination due to spills is "a matter of when,
not if." However, Michigan DEQ officials told us that several factors make
it unlikely that spills in the St. Clair River will contaminate drinking
water:

           o  Drinking water intakes are 20-30 feet below the water's
           surface.

           o  The river has distinct channels, and it is difficult for a
           pollutant originating on one side of the river to cross these
           channels.

           o  At 180,000 to 200,000 cubic feet per second, the river flows so
           quickly that pollutants are flushed downstream before they affect
           drinking water.

           In contrast, Michigan DEQ officials told us that Canadian drinking
           water facilities are more vulnerable to contamination from spills
           in the St. Clair River. These officials noted that Canadian
           drinking water facilities have shut down more often than Michigan
           facilities as a result of spills in the corridor. They noted that
           the most vulnerable Canadian drinking water facility is located on
           Walpole Island, directly downstream of Sarnia, and it provides
           drinking water to members of a First Nation community.12

           Currently there are, or will soon be, efforts under way to
           supplement the existing spill notification processes employed by
           the U.S. and Canadian Coast Guards, and Michigan and Ontario
           officials. Informal notification processes are already being
           employed along the corridor. For example, a local emergency
           management coordinator in the St. Clair River area of the corridor
           has developed an informal agreement with Canadian industry
           representatives to call and notify him directly of any spills into
           the St. Clair River. Upon receiving spill information, he provides
           the information directly to drinking water facility operators
           along the portion of the corridor that borders the St. Clair
           River. Three drinking water facility operators listed him as their
           first source of spill information. In addition, Sarnia-Lambton
           Environmental Association (SLEA) officials told us that their
           member facilities contact Michigan drinking water facilities
           directly in the event of a spill. Several drinking water facility
           operators confirmed that they have received notification from
           members of this consortium of Canadian industries.

           In addition, two monitoring systems are being developed by
           officials from counties bordering the corridor and Michigan DEQ
           officials, who have obtained federal grants to install spill
           detection equipment in the St. Clair and Detroit Rivers. These
           systems are designed to provide spill information directly to
           drinking water facility operators with water monitoring equipment
           located near their intakes. One monitoring system, for the St.
           Clair River and Lake St. Clair, is funded by an EPA grant of
           $962,200 to Macomb and St. Clair Counties. The other monitoring
           system, for the Detroit River, is funded by a DHS grant of
           $760,000 to Michigan DEQ. The officials involved in obtaining both
           grants told us they are coordinating their efforts so that an
           overall network of water quality monitors will be more seamless
           along the corridor. For example, they plan to purchase the same
           monitoring equipment so that maintenance can be shared. EPA and
           Michigan DEQ officials estimate that the monitoring systems will
           be in place in the St. Clair and Detroit Rivers no later than 2007
           (see fig. 8). These systems are based on the Ohio River Valley
           Sanitation Commission's (ORSANCO) spill detection and notification
           system, established in 1978 to protect drinking water intakes from
           chemical contamination. For additional information on this system,
           see appendix VI.

           Figure 8: Existing and Proposed Water Quality Monitoring Stations
           in the St. Clair-Detroit River Corridor

           EPA's spill prevention program addresses only oil spills, and EPA
           is uncertain as to which facilities are governed by its spill
           prevention requirements. EPA Region 5 conducted varying numbers of
           spill prevention-related inspections per year in the corridor for
           the time frame we reviewed, and their inspections uncovered
           significant spill prevention deficiencies. In contrast, the Coast
           Guard's spill prevention efforts include oil and hazardous
           substances. The Coast Guard's District 9 inspections targeted a
           greater number of the facilities and vessels they regulate;
           however, the Coast Guard's inspections were multi-mission rather
           than focused on spill prevention exclusively. Their inspections
           revealed minor spill prevention-related issues. In response to
           spills and noncompliance issues, EPA and the Coast Guard issued a
           total of 16 penalties in the time period we reviewed.

           While EPA has the authority to address spill prevention for both
           oil and hazardous substances, its program only addresses oil. In
           1972, in amendments to the Clean Water Act, Congress called for
           regulations to prevent discharges of oil and hazardous substances;
           in 1974, EPA's SPCC program became effective. EPA's regulations
           require non-transportation related facilities with specified oil
           storage capacities13 which, because of their location, could
           reasonably be expected to discharge oil into the navigable waters,
           to implement a SPCC plan that has been certified by a licensed
           engineer. These plans should identify the location and types of
           stored oil, discharge prevention measures, drainage controls, and
           methods of disposal. Facilities must also meet certain operational
           standards that include having

           o  necessary containment structures or equipment;14

           o  periodic integrity tests of containers and leak tests of valves
           and piping; and

           o  training for oil-handling personnel on equipment operation and
           maintenance, discharge procedure protocols, pollution control laws
           and rules, facility operations, and the contents of their
           facility's SPCC plan.

           In the late 1970s, EPA proposed hazardous substance spill
           prevention regulations, but they were never finalized. EPA
           officials speculated that, possibly, these regulations were not
           finalized because oil spills were more prevalent, hazardous
           substance spills have shorter-term effects than oil spills, and
           because EPA focused on the NPDES program to control chronic
           pollutant discharges.

           While EPA's spill prevention program targets oil spills, the Coast
           Guard's program addresses spill prevention for both oil and
           hazardous substances. The program applies to facilities or vessels
           that are capable of transferring oil or hazardous materials, in
           bulk, to or from vessels of certain minimum capacity. Facilities
           are required to develop an operations manual, employ qualified
           personnel, and meet equipment standards. The operations manual
           must contain a description of the facility layout, the location of
           important equipment and personnel, and a discussion of procedures
           for transfer operations and emergencies. The manual must also
           include a summary of applicable laws and information concerning
           personnel training and qualifications. Also, each facility must
           have emergency shutdown capacity and specified discharge
           containment features. Vessels are required to have written
           transfer procedures for oil and hazardous substances, meet
           maintenance and equipment standards, and employ qualified
           personnel.

           In addition to the Coast Guard and EPA's prevention programs, the
           Michigan DEQ has a spill prevention program that is administered
           in conjunction with their NPDES program. This program requires
           that facilities that store or use oil or polluting substances, or
           those that may be deemed a hazard to waters of the state, create
           and implement spill prevention plans and inform Michigan DEQ of
           the plan's completion and availability upon request.15 Michigan
           DEQ's pollution prevention plans are to include a

           o  polluting material inventory;

           o  detailed facility plan, including floor drains and loading
           areas;

           o  secondary storage container description; and

           o  discussion of precipitation management.

           The plans are also to include spill control and cleanup procedures
           and are required to be reevaluated every 3 years (or whenever a
           material release occurs). If a facility is also subject to EPA's
           SPCC program, it may submit a combination spill prevention plan
           that meets both state and federal requirements. If the facility is
           only subject to the Michigan DEQ's spill prevention planning
           requirements, it is not required to have its plans certified by an
           engineer.

           On the Canadian side of the St. Clair-Detroit River Corridor, the
           Ontario Ministry of Environment did not have spill prevention
           regulations in place in the time frame we reviewed. Instead, the
           Ministry issued orders which required individual companies to
           conduct spill prevention planning, or it required spill prevention
           planning as a requirement for companies seeking a Certificate of
           Approval, which is required before operating. Due in part to the
           large chemical spills in 2003 and 2004 originating from facilities
           in Sarnia, the Ontario Ministry of Environment introduced new
           legislation under its Environmental Protection Act which addresses
           the requirement for spill prevention planning.16

           EPA Region 5 does not know the universe of facilities that are
           subject to its spill prevention program requirements and it
           conducts varying numbers of inspections of known facilities under
           its jurisdiction in the corridor. Facilities that must comply with
           SPCC regulations are not required to report to the agency, so EPA
           does not have an inventory of facilities it regulates. The
           challenge this presents is not limited to the corridor, as EPA
           officials are uncertain as to how many facilities should comply
           with SPCC program requirements nationwide. In the corridor, EPA
           Region 5 has identified 59 facilities (of a greater number) that
           are required to meet SPCC requirements, either through special
           multi-media inspection initiatives or by referrals from Michigan
           DEQ.

           While SPCC plans must be reevaluated and reviewed every 5 years, a
           specified inspection frequency is not contained in EPA's
           regulations. EPA officials in Region 5, which encompasses the
           corridor, rely on roughly three SPCC inspectors to conduct all
           plan reviews and provide all compliance assistance for facilities
           in the six-state region. According to these officials, with
           current SPCC resource constraints, they could only inspect
           facilities once every 500 years or more. From 1994 to 2004, EPA
           Region 5 inspected an average of 10 percent of the 59 known
           SPCC-regulated facilities in the corridor per year. (See fig. 9.)

           Figure 9: Number of SPCC Inspections Conducted in the St.
           Clair-Detroit River Corridor from 1994-2004

           Note: In 2001, EPA Region 5 did not conduct any SPCC inspections
           in the corridor.

           EPA Region 5 inspected a number of these SPCC-regulated facilities
           as part of several multi-media inspection efforts conducted by
           their Enforcement and Compliance Assistance Team, including the
           Detroit River and Flyway Enforcement and Compliance Assistance
           Initiative. This effort identified and inspected 28 facilities in
           the Detroit area for compliance with multiple EPA programs,
           including the SPCC program; some of the inspected facilities
           overlap with a portion of the facilities along the corridor. When
           SPCC program officials inspect a facility they use a standardized
           approach, which includes the following:

           o  an in-depth review of the facility's SPCC plan;

           o  an interview with the facility owner or operator;

           o  a physical inspection of the facility;

           o  a verification of equipment, containment structures, and
           buildings;

           o  a review of facility inspections and training records; security
           and integrity testing; and

           o  verification that the facility's SPCC plan has been certified
           by a licensed engineer.

           While EPA has a separate program for spill prevention, the Coast
           Guard addresses spill prevention during its routine safety and
           security inspections of facilities and vessels. The Coast Guard's
           District 9 regulates over 100 facilities and 23 vessels stationed
           in the corridor, as well as vessels that travel through the
           corridor. It also regulates the transfer of oil and hazardous
           substances. The Coast Guard inspects facilities and vessels to a
           much greater extent per year than EPA; however, its inspections
           are multi-purpose rather than focused exclusively on spill
           prevention. The Coast Guard's annual facility inspections
           incorporate spill prevention components that are similar to EPA's
           SPCC inspection components, but their material transfer
           inspections and spot checks are not comparable to EPA's focused
           spill prevention inspections. From 1994 to 2004, the Coast Guard's
           District 9 inspected an average of 44 facilities, 135 vessels, and
           30 material transfer events per year, for safety, security, and
           pollution prevention requirements. When material transfer events
           are excluded, the Coast Guard inspected, on average, about 44
           percent of the facilities in their jurisdiction per year-compared
           to EPA's inspections of roughly 10 percent of the known
           SPCC-regulated facilities. However, we are uncertain as to how
           many of the Coast Guard's yearly inspections were on-site
           inspections that are comparable to EPA's SPCC inspections as
           opposed to spot checks or other multi-purpose inspections that the
           Coast Guard conducts. The Coast Guard conducts regular on-site
           inspections that consist of a check of maintenance and operation
           procedures, as well as a facility or vessel's spill prevention
           planning. For their annual facility inspections, Coast Guard
           officials review, among other items:

           o  contents of operations manuals, including specifications for
           containment equipment;

           o  transfer equipment requirements, including an examination of
           transfer pipes for defects; and

           o  facility operations, including whether the designated person in
           charge has certification of completion of required training.

           While some inspections are conducted on-site, the Coast Guard also
           conducts remote examinations, such as viewing a transfer of
           materials from a distance using binoculars. The specific type and
           number of inspections conducted by the Coast Guard from 1994
           through 2004 is shown in table 2.

           Table 2: Number of Coast Guard Facility, Vessel, and Transfer
           Inspections Conducted in the St. Clair-Detroit River Corridor from
           1994-2004

           Source: GAO analysis of the Coast Guard's data.

           aData were not provided for 2004.

           In addition to EPA and the Coast Guard, the Michigan DEQ inspects
           facilities for compliance with its spill prevention program during
           its regular NPDES program inspections. According to Michigan DEQ
           officials, their inspectors do not keep track of the number of
           spill prevention inspections conducted or deficiencies found due
           to a lack of funding for its spill prevention program. Further,
           the universe of facilities regulated by its spill prevention
           program is unknown-but approximately 400 facilities that are in
           Michigan DEQ's Southeast District, which is a larger area that
           includes the U.S. side of the corridor-have submitted certified
           spill prevention plans. On the Canadian side of the corridor,
           Ontario's Ministry of Environment conducts inspections that
           include a spill prevention component. Ministry of Environment
           officials were able to provide inspection data from 2003 to 2005,
           which indicated that they inspected roughly 35 petrochemical and
           related facilities per year. The inspections conducted in 2004 and
           2005 reflect the work that an "Environmental SWAT Team"
           conducted.17 The focus of this special initiative was on
           facilities with the potential for future spills that could pose
           risks to human health and the environment. The inspections
           included a comprehensive review of the facilities' air emissions,
           water discharges, and spill prevention and contingency plans.

           EPA Region 5 officials told us that the spill prevention
           inspections they conducted from 1994 through 2004 disclosed
           significant and numerous deficiencies, such as failure to provide
           for secondary containment or failure to prepare spill prevention
           plans. For example, an SPCC inspector found that one company
           failed to prepare its SPCC plan within 6 months of beginning
           operations, and failed to implement its plan within 1 year. In
           addition, the facility never had its SPCC plan certified by a
           professional engineer. The SPCC inspector found that another
           facility had no additional containment around some bulk storage
           tanks, and it failed to amend its SPCC plan as required. In
           contrast, Coast Guard officials from District 9 told us that their
           inspections revealed that nearly all facilities and vessels were
           in compliance, and those that were not had only minor
           noncompliance issues related to spill prevention, such as
           incidental omissions in operations manuals. For example, Coast
           Guard officials found instances in which a facility operator
           initialed only one section of a required form, rather than at
           multiple sections. The Coast Guard officials also found other
           minor violations that related to aged hoses outside their service
           life and inadequate lighting.

           Michigan DEQ officials told us that their inspections revealed
           that some facilities do not have spill prevention plans, or did
           not certify compliance with them. In some cases, facilities that
           already had some secondary containment or protection in place
           needed further upgrades in order to come into compliance with the
           state's spill prevention requirements. For example, during one
           inspection, Michigan DEQ found that a facility had developed a
           spill prevention plan, but it was not adequately managing its
           materials in order to prevent storm water from contacting the
           materials and discharging them into the waterways. On the Canadian
           side, Ontario's Environmental SWAT Team found that 34 of the 35
           facilities inspected in 2004 and 2005 were not in compliance with
           one or more legislative and regulatory requirements. Eight
           facilities did not have spill prevention and contingency plans.
           Other deficiencies found during inspections included:

           o  a Certificate of Approval was not obtained for operations;

           o  equipment, systems, processes, or structures were altered
           contrary to the existing Certificate of Approval; and

           o  chemicals were improperly handled, stored, and identified.

           EPA and the Coast Guard issued 16 penalties in response to spills,
           noncompliance with spill prevention programs, or for failure to
           report spills during the period we reviewed. EPA Region 5 issued
           four penalties, primarily for SPCC violations, that were assessed
           at an average of $39,000 each during the 11-year period. During
           the same time period, the Coast Guard's District 9 issued 12
           penalties for spills that were assessed at an average of $2,100
           each. See figure 10 for total amounts of penalties assessed by EPA
           and the Coast Guard per year from 1994 through 2004.

           Figure 10: EPA and Coast Guard Spill-Related Penalties Assessed
           from 1994 to 2004 in the St. Clair-Detroit River Corridor

           Note: In 1998, 2000, 2002, and 2004 no financial penalties were
           issued by EPA or the Coast Guard.

           EPA Region 5 officials told us they rely primarily on assisting
           companies in coming into compliance with spill prevention program
           regulations, and they pursue enforcement actions and issue
           financial penalties when companies fail to respond to their
           assistance. They also explained that limited staff resources are
           available to pursue enforcement actions. EPA Region 5 has the
           equivalent of roughly one and a half full time staff persons
           devoted to spill-related enforcement duties conducted by the
           Office of Regional Counsel and the Oil Program. The Oil Program is
           responsible for determining noncompliance with the SPCC program,
           which entails establishing a history of spills or noncompliance;
           it is then responsible for determining penalty amounts.

           Determining noncompliance and identifying responsible parties for
           spills, however, can be problematic for EPA. While EPA Region 5
           officials told us they use their spill data when pursuing an
           enforcement case, the agency, for most spills, does not confirm
           the validity of the spill data or gather additional information.
           Michigan DEQ responds to most reported spills in the corridor, but
           EPA Region 5 does not coordinate with Michigan DEQ to collect
           information for enforcement purposes. It is EPA's policy to
           collect spill information directly because, according to EPA
           officials, it is preferable to have first-hand knowledge in the
           event that staff have to testify or provide a deposition for an
           enforcement case, among other reasons. To gather additional
           information, Oil Program officials stated that they send requests
           to facilities for spill-related information that they can then use
           in enforcement cases. If spill information is obtained, EPA Region
           5 officials told us that their informal policy is not to pursue an
           enforcement case when the proposed penalty is less than $11,000 or
           the spill involves less than 100 gallons or two barrels of oil.
           They also stated that in most cases for which they issued a
           penalty, the amounts were ultimately reduced substantially through
           negotiations with the responsible party. For example, one facility
           was issued a financial penalty of approximately $320,000 for a
           spill prevention violation and the negotiated final payment was
           $25,000.

           Like EPA, the Coast Guard relies primarily on assisting companies
           in coming into compliance with spill prevention program
           regulations. The average financial penalties that were assessed by
           the Coast Guard's District 9 were relatively low compared to the
           maximum financial penalty of $32,500 that it has the authority to
           issue for a spill violation. Coast Guard officials from District 9
           told us that, in determining penalties, they take into account how
           much a facility or vessel owner has already paid for the cost of
           cleaning up a spill.18 In regard to why no penalties were assessed
           for spill prevention program violations, officials from the Coast
           Guard stated that they have the authority to order a facility or
           vessel to cease operations if it does not comply with their spill
           prevention program, and this serves as a strong deterrent to
           noncompliance. They added that large financial penalties have not
           been needed due to the cooperation of companies coming into
           compliance with their prevention regulations.

           Similar to the Coast Guard, Michigan DEQ did not issue penalties
           for noncompliance with their spill prevention program in the time
           frame we reviewed. Michigan DEQ did, however, issue four penalties
           averaging $35,000 to responsible parties for spills in the
           corridor. Michigan DEQ also issued three penalties for multiple
           violations including spills and industrial permit violations; the
           penalties totaled approximately $300,000. Lastly, in Ontario, the
           Ministry of Environment did not have the authority to issue
           administrative penalties for spills until 2005.19 Prior to that
           time, the Ministry of Environment pursued spill-related penalties
           through the provincial court system and from 2002 to 2004, four
           facilities were assessed penalties averaging approximately
           $171,000 in U.S. dollars after successful prosecutions.

           In addition, though EPA and Coast Guard officials acknowledge that
           spill reporting is not always occurring, the agencies did not
           penalize responsible parties for failure to report chemical
           releases in the time period we reviewed. We are, however,
           uncertain as to how many chemical releases occurred in the
           corridor due to data limitations. The authority provided under the
           EPCRA allows EPA officials to penalize regulated industries for
           failure to report, in a timely fashion, spills of reportable
           quantities of hazardous or extremely hazardous substances.
           However, EPA Region 5's Chemical Preparedness officials who
           administer EPCRA told us that they rely on 30 to 40 information
           requests sent to companies per year for the entire six-state
           region to gather the necessary information on reporting in order
           to pursue enforcement. They stated that, with three staff for the
           region to enforce the EPCRA and other regulations related to
           chemical releases, they lack the resources to inspect more than
           roughly 15 facilities per year to determine compliance with timely
           notification and other reporting requirements. They did not issue
           any penalties for failure to report chemical releases in the
           corridor in the time frame we reviewed. Superfund authorizes EPA
           and the Coast Guard to issue penalties for failure to report
           hazardous substance spills, but neither agency did so in the time
           frame we reviewed. The Clean Water Act, on the other hand, does
           not authorize civil penalties for a responsible party's failure to
           notify NRC of an oil spill; only criminal sanctions are available.
           Michigan DEQ officials may make a criminal complaint or request
           that the state's Attorney General pursue civil action for failure
           to report spills; however, they did not in the period we reviewed.
           On the Canadian side of the corridor, Ontario's Environmental
           Protection Act provides Ministry of Environment officials with the
           authority to penalize responsible parties for failure to report
           spills, as this is a violation of the act. The Ministry did so
           recently, and the company responsible for a large chemical spill
           into the St. Clair River in 2003 was charged and convicted of
           failing to report the spill immediately.

           Spill-related penalties that are collected by EPA and the Coast
           Guard help supplement the funds that provide for response efforts
           when a responsible party is not identified to pay costs. In these
           cases, EPA and the Coast Guard may obtain financing from the Oil
           Fund or Superfund to pay for their response efforts, including the
           cleanup.20 But these funds are being depleted by cleanup efforts
           in the corridor and are not being replenished through the cost
           recovery process because, in many cases, the responsible parties
           have not been identified. Fund data maintained by the National
           Pollution Funds Center show that, from 1994 to 2004,

           o  approximately $8.4 million21 from the Oil Fund financed oil
           spill cleanups in the corridor, for which $80,067 was recovered to
           offset those expenditures, and

           o  approximately $17,000 from Superfund financed hazardous
           material spill cleanups, with no additional funds recovered to
           offset those expenditures.

           Spills of oil and hazardous substances into waters of the corridor
           continue to be a concern, and agencies responsible for addressing
           this problem face challenges on several fronts including obtaining
           accurate spill information, and conducting spill notification and
           comprehensive prevention efforts. Officials from EPA Region 5 and
           Coast Guard's District 9 concur that accurate spill information is
           not available and acknowledge that such information could be
           improved by better incorporating data, including final spill
           volume estimates, obtained through their spill response efforts.
           Coast Guard officials from District 9 also acknowledge that they
           could help update the NRC's spill information by documenting which
           duplicative NRC spill reports are linked to common incidents.
           Better documentation of response efforts and the results of spill
           investigations could assist EPA and the Coast Guard in targeting
           inspection and enforcement efforts to the highest-priority need.
           Spill notification under the agreement between the United States
           and Canada, and Michigan and Ontario, while limited, appears to be
           meeting its intended purpose. Effective spill prevention helps
           reduce contaminants flowing into waters of the corridor. EPA's
           spill prevention efforts are hampered by the fact that it does not
           know the universe of facilities regulated by its program, the
           scope of its program is focused only on oil, and limited resources
           are available for implementing the program-particularly for
           inspections. Given the focus and resource limitations which impact
           EPA's ability to pursue spill prevention and enforcement
           activities, EPA could collect information about the facilities
           that are regulated in order to better define goals for the
           frequency and extensiveness of their inspections.

           To better ensure that spill data are available to target their
           inspection and enforcement efforts, and to improve the overall
           effectiveness of spill notification, we are recommending that the
           EPA Administrator direct EPA Region 5; and that the Secretary of
           Homeland Security direct the Commandant of the Coast Guard and the
           Commander of District 9 to take the following two actions:

           o  maintain and update spill information to include the results of
           investigations and explore the feasibility of updating spill
           information maintained by the NRC, and

           o  determine whether existing spill notification processes can be
           improved or modified to provide reduced and consistent
           notification time frames.

           In addition, to better utilize spill prevention resources, we
           recommend that the EPA Administrator consider gathering
           information on which facilities are regulated under its spill
           prevention program. We also recommend that the EPA Administrator
           direct Region 5 to develop goals for the frequency and
           extensiveness of its inspections.

           GAO provided a draft of this report to EPA and DHS for review and
           comment. DHS provided comments on the draft report and generally
           agreed with our findings and conclusions. EPA provided only
           technical comments regarding the report. DHS did not address our
           recommendations. EPA commented on the feasibility of our
           recommendations regarding gathering information on SPCC regulated
           facilities and updating spill information maintained by the NRC.

           While DHS generally agreed with our findings and conclusions, the
           agency commented on our observation that the Coast Guard does not
           update spill volume estimates in its automated spill data system.
           Specifically, the agency cited an example used in our report and
           noted that it was the result of unusual circumstances that arose
           during the transition from one data system to another. In
           addition, DHS noted that Coast Guard investigators do have the
           ability to update spill volume estimates in investigative report
           narratives. We acknowledge that the Coast Guard transitioned from
           one data system to another in the time frame we reviewed. However,
           Coast Guard officials told us that with the current data system,
           initial spill volume estimates cannot be readily updated, except
           in the narrative of the investigation reports. We acknowledge that
           investigators do have the ability to update volumes in the report
           narratives but initial volume estimates cannot be changed in the
           volume data field of the current system. It is difficult to
           readily assess the magnitude of spills based on the initial volume
           estimates contained in the automated spill data. Furthermore,
           Coast Guard officials told us that they would benefit from an
           additional field in their spill data system which incorporated
           final spill volume estimates. In addition, DHS commented that
           additional factors should be considered in our report regarding
           agency efforts to penalize responsible parties for failure to
           report spills. DHS acknowledged that the responsible parties for
           many spills are not identified and, while penalties were not
           assessed for failure to report spills in the corridor in the time
           frame we reviewed, DHS stated that the Coast Guard, with the
           Department of Justice, has successfully prosecuted responsible
           parties for spills outside the corridor. The full text of DHS's
           comments is included in appendix VII.

           EPA provided the following three comments on our report. First,
           EPA stated that it does respond to every spill, whether directly
           or indirectly, in the same way that the Coast Guard responds.
           However, EPA could not provide documentation of its response
           efforts whereas the Coast Guard provided documentation that
           indicated what actions were taken in response to each spill. In
           addition, EPA stated that its spill responders rely heavily on
           Michigan DEQ to respond to spills and coordinate response actions
           with the Coast Guard. EPA also said that it responded directly to
           greater than one percent of spills-as opposed to less than one
           percent, as previously stated by EPA officials. Nevertheless, the
           level of EPA's response is unclear due to lack of documentation.
           Second, EPA commented on our recommendation that it update spill
           information maintained by the NRC. We acknowledge that EPA does
           not modify spill data maintained by the NRC; however, our
           recommendation was that it explore the feasibility of updating
           spill information maintained by the NRC by informing NRC of
           duplicate spill reports. While EPA maintains that information on
           spills can now be updated using a new system called Web EOC, and
           our report acknowledges that Web EOC is a method to update spill
           data electronically, the extent of its use is uncertain according
           to EPA officials. Finally, EPA commented on the feasibility of our
           recommendation that it gather information on facilities that are
           covered under its spill prevention program. EPA stated that there
           is no authority in the Clean Water Act or the prevention
           regulations for facilities to provide this information to EPA. It
           further stated that under the Paperwork Reduction Act the agency
           would need to seek approval from the Office of Management and
           Budget. However, EPA has previously identified SPCC facilities in
           the corridor. If the agency determines that formal rulemaking is
           necessary for it to gather information on which facilities are
           covered under its spill prevention program, then we believe it
           should consider undertaking such a rulemaking. EPA officials also
           provided specific technical comments and clarifications on the
           draft report that we have incorporated in the report as
           appropriate.

           As agreed with your offices, unless you publicly announce the
           contents of this report earlier, we plan no further distribution
           until 30 days from the report date. At that time, we will send
           copies of this report to appropriate Congressional Committees, the
           EPA Administrator, the Secretary of Homeland Security, and various
           other federal and state agencies. We also will make copies
           available to others upon request. In addition, the report will be
           available at no charge on the GAO Web site at http://www.gao.gov .

           If you or your staff have any questions, please call me at (202)
           512-3841 or [email protected]. Contact points for our Offices of
           Congressional Relations and Public Affairs may be found on the
           last page of this report. GAO staff who made major contributions
           to this report are listed in appendix VIII.

           John B. Stephenson Director, Natural Resources and Environment

           List of Congressional Requesters

           The Honorable Carl Levin United States Senate

           The Honorable Debbie Stabenow United States Senate

           The Honorable Sander M. Levin House of Representatives

           We were asked to examine (1) how many oil and hazardous substance
           spills of more than 50 gallons (or of an unknown volume) were
           reported in the St. Clair-Detroit River corridor from 1994 to
           2004, and how accurately reported spills reflect the extent of
           actual spills; (2) what processes are used to notify parties of
           spills, and whether they contain explicit requirements for
           reporting times and spill magnitude; and (3) the extent of
           Environmental Protection Agency (EPA) and the Coast Guard's spill
           prevention efforts and enforcement activities in the St.
           Clair-Detroit River corridor from 1994 through 2004.

           To determine the number of oil and hazardous substance spills of
           more than 50 gallons (or of an unknown volume) reported in the St.
           Clair-Detroit River corridor from 1994 to 2004, and to what extent
           they represent actual spills, we obtained information on spills
           with those characteristics reported in the St. Clair River, Lake
           St. Clair, the Detroit River, and a highly industrialized
           tributary, the Rouge River. We obtained data sets with these
           attributes from the National Response Center (NRC), EPA Region 5,
           the Coast Guard's Headquarters, Michigan's Department of
           Environmental Quality (Michigan DEQ), and the Spills Action Centre
           (SAC) operated by the Ontario Ministry of Environment. To assess
           the reliability of each data set we questioned and interviewed
           knowledgeable officials about the data and the systems that
           produced them, and manually reviewed the data. Limitations to the
           data are discussed in the report and in appendix II. When
           appropriate, we analyzed the data sets individually to determine
           spill frequency over time and spill characteristics, such as
           volume and the type of material spilled. We were not able to
           combine the spill data sets for analysis because each entity
           tracks spills differently, and we were limited in what we could
           conclude from the individual data sets because the degree to which
           they are updated varies widely. We also obtained EPA and Michigan
           DEQ data sets related to other pollutant discharges, such as
           combined sewer overflows (CSO) and industrial discharge permit
           violations, to provide context for the spill data and obtain more
           complete information on pollutants discharged into the water
           bodies of the St. Clair-Detroit River corridor. These data are
           likely subject to the same limitations as the spill data, in that
           industrial permit violations and CSOs are self-reported and
           facilities may be reluctant to report these events; however,
           spills may be particularly subject to underreporting because they
           are not part of a structured program as are CSOs and industrial
           permit violations.

           To assess what processes are used to notify parties of spills and
           whether they contain explicit requirements for reporting times and
           spill magnitude, we reviewed applicable laws and spill
           notification agreements and obtained information on implementation
           of these agreements from EPA, the Coast Guard, Michigan DEQ, the
           Michigan State Police, and Canadian officials. In addition, we
           obtained and analyzed documentation on six spills to better
           understand how the notification process was conducted in specific
           incidents. We selectively sampled the spill data sets for spills
           to illustrate the implementation of notification practices under
           various scenarios, including spills with source locations in both
           the United States and Canada and spills of differing materials and
           volumes. We questioned the 17 drinking water facility operators on
           the U.S. side of the corridor to obtain their perspectives on the
           timeliness of spill notification. We further obtained information
           on the automated monitoring system maintained by Sarnia-Lambton
           Environmental Association, planned automated monitoring on the
           U.S. side of the corridor, and the monitoring conducted by the
           Ohio River Valley Water Sanitation Commission.

           To determine the extent of EPA and the Coast Guard's spill
           prevention efforts and enforcement activities in the St.
           Clair-Detroit River corridor from 1994-2004, we first obtained and
           analyzed laws, regulations, and agency policies regarding spill
           prevention and enforcement. This included obtaining information on
           the potential enforcement penalty dollar amounts. We also obtained
           data from EPA, the Coast Guard, Michigan DEQ, and the Ontario
           Ministry of Environment on spill-related enforcement actions taken
           in the corridor since 1994. We analyzed the information to
           determine the number of inspections conducted, the types of
           violations found, and the penalties assessed for each documented
           violation. Finally, we obtained information from the various
           agencies on the resources devoted to inspections and enforcement;
           the use of those resources; and priorities in using the resources.

           We performed our work from September 2005 to June 2006 in
           accordance with generally accepted government auditing standards.

           Spill data sets were available from four sources: EPA, the Coast
           Guard, Michigan DEQ, and the Ontario SAC. Each data set is unique;
           however, some spill incidents are found in multiple data sets and
           therefore they cannot be combined or consolidated. The relative
           quality of each data set depends in part on whether it is updated
           after additional information is obtained from spill investigations
           or whether minimal updates are made, as with the EPA spill data
           set. Generally, all of the spill data sets have a common data
           reliability limitation which stems from uncertainty regarding
           whether all incidents are reported. Of note, the data sets for EPA
           and the Ontario SAC contained a large number of incidents with
           unknown volumes.

           EPA's spill data set is not routinely updated after EPA responders
           conduct investigations. Therefore, the data reflect preliminary
           information about spills received from the NRC, and the data
           likely do not represent the actual number and nature of spills. We
           are presenting this data for informational purposes only. The data
           set contained a total of 916 spill incidents that occurred in the
           St. Clair River, Lake St. Clair, Detroit River, and Rouge River
           from 1994 through 2004 and that had volumes of greater than 50
           gallons (or of an unknown volume). About 45 percent of the spills
           were oil-related. The number of spills has varied over time, not
           showing either an increasing or decreasing trend. The EPA data
           showed that the greatest number of spills occurred in 1994.

           Table 3: EPA Spill Data by Year and Material

           Source: GAO analysis of EPA data.

           aSpills in the oils category consist of water with oil (2),
           miscellaneous waste oil (1), hydraulic oil (39), fuel oil (68),
           miscellaneous oil (152), diesel oil (74), other oil (34),
           transmission oil (3), waste oil (23), turbine oil (2), edible oils
           (4), epoxidized vegetable oils (2), capacitor oil (1), petroleum
           products (1), crude oil (3), compressor oil (1), engine oil (1),
           sludge oil (1), carbon oil (1), and crank oil (1).

           bSpills in the gasoline and fuels category consist of automobile
           gasoline (81), jet fuel (20), kerosene (3), unleaded gasoline (3),
           gasoline or diesel (1), bunker c fuel oil (2), and fuel and
           lubricant (1).

           cSpills in the chemicals category consist of mineral spirits (4),
           other chemicals (1), solvents (2), paints (2), Anton battery
           release (1), and hydrocarbon toluene (1).

           dSpills in the tars category consist of coal tar pitch (3), crude
           coke oven tar (1), and anthracene oil (1).

           eSpills in the wastewater category consist of wastewater with
           vegetable oil (1).

           fSpills in the other category consist of brake fluid (1), asphalt
           (3), parking lot runoff (1), industrial waste (1), white foam
           material (1), and bilge slops (1).

           Coast Guard officials update spill data after investigations are
           conducted, thereby strengthening the reliability of their spill
           data. However, they are unable to update preliminary volume
           estimates and therefore these data are likely unreliable. There
           are 51 spill incidents in the Coast Guard data set and the
           majority of spills-roughly 70 percent-were oil-related. The Coast
           Guard's spill data set indicates that 11 spills were traced back
           to storm or sanitary sewer outfalls. In four of these instances,
           narratives completed by spill responders indicate that sewage was
           mixed with other spill materials.

           Table 4: Coast Guard Spill Data by Year and Material

           Source: GAO analysis of Coast Guard data.

           aAsphalt.

           bJet fuel.

           The Coast Guard's data shows that the greatest number (26 of the
           51) spills occurred in the Detroit River. Most of the oil spills
           investigated by the Coast Guard were in the Detroit and Rouge
           Rivers. Similarly, most of the chemical spills that the Coast
           Guard investigated were in the Detroit River, while most of the
           gasoline spills were in Lake St. Clair.

           Table 5: Coast Guard Spill Data by Water Body and Material

           Source: GAO analysis of Coast Guard data.

           aOne of which was jet fuel.

           bAsphalt.

           cOne spill incident did not have a water body listed.

           Michigan DEQ officials update their spill data after
           investigations are conducted, but some data fields (e.g., quantity
           of material released) are not completed because the information is
           unknown. There are 21 spill incidents in the Michigan DEQ spill
           data set that occurred in the St. Clair River, Lake St. Clair,
           Detroit River, and Rouge River from 1996 through 2004 that have
           volumes of greater than 50 gallons (or of an unknown volume).
           Michigan DEQ did not provide spills prior to 1996 because that is
           when they began collecting spill data electronically.

           Table 6: Michigan DEQ Spill Data by Year and Material

           Source: GAO analysis of Michigan DEQ data.

           aSpills in the chemicals category consist of sodium bisulfate (1),
           bromine (1), hydro fluranosilic acid (1), and sodium hydroxide
           (1).

           bSpills in the oils category consist of non-PCB oil (1) and oils
           (4).

           cSpills in the wastewater category consist of untreated wastewater
           (1), phosphorus wastewater (1), contact wastewater (1), and
           untreated process wastewater (1).

           dSpills in the gasoline and fuels category consist of diesel (2).

           eSpills in the tars category consist of coal tars (2).

           fSpills in the other category consist of brine (2), paper coatings
           (1), and clay coatings (1).

           Table 7: Michigan DEQ Spill Data by Water Body

           Source: GAO analysis of Michigan DEQ data.

           Ontario Ministry of Environment officials update spill data to
           reflect additional information obtained. However, not all data
           fields are completed because information such as spill quantities
           and materials are not always known. There are a total of 157 spill
           incidents in the SAC data that occurred in the St. Clair River
           (105), Lake St. Clair (5), and the Detroit River (47) between 1994
           and 2004 that have volumes greater than 50 gallons (or of an
           unknown volume). About 9 percent of the 157 have unknown
           responsible parties, and 127 of the 157 have unknown volumes or
           masses.

           Table 8: Ontario SAC Spill Data by Year and Material

           Source: GAO analysis of Ontario SAC data.

           aSpills in the chemicals category consist of acetonitrile,
           ammonia, benzene, butylene, cyclohexane, divinylbenzene, epoxy,
           ethylene, ethyl hexyl nitrate, ethyl benzene, ferric chloride,
           iron, metabisulphate, methylene chloride, methyl ethyl ketone,
           n-hexane, propylene dichloride, rust inhibitor, sodium chloride,
           styrene monomer, toluene, vcm, and xylene.

           bSpills in the oils category consist of heating oil, oily
           material, transformer oil, petroleum oil, vegetable oil, and waste
           oil.

           cSpills in the wastewater category consist of ash and water,
           chlorinated water, contaminated water, wastewater, wastewater with
           aromatics, wastewater with chlorides, wastewater with
           hydrocarbons, process water, effluent water, and water.

           dSpills in the gasoline and fuels category consist of diesel fuel,
           fuel, and gasoline.

           eSpills in the unknown category consist of unknown, not
           applicable, and blank fields.

           fSpills in the other category consist of asphalt, brine, cement,
           concrete, cooling water, foam, plastic material, sand and gravel,
           sediment, silt, and suspended solids.

           Table 9: Ontario SAC Spill Data by Water Body

           Source: GAO analysis of Ontario SAC data.

           Michigan DEQ's CSO data were available as of 1999, when Michigan
           DEQ began tracking sewer overflows. The CSO data, like spill data,
           have a data-reliability limitation relating to uncertainty as to
           whether all CSO events are reported; however, spills may be
           particularly subject to underreporting because they are not part
           of a structured program as CSOs and industrial permit violations
           are. CSO data provide additional information in terms of the
           amount and location of pollutant discharges into the waters of the
           corridor. According to EPA, CSOs contain storm water, untreated
           human and industrial waste, toxic materials, and debris. The
           roughly 1,400 CSOs that were greater than 50 gallons (or of an
           unknown volume) greatly exceeded the number of spills that met
           these criteria during the 6-year period. The largest category of
           CSOs was of diluted raw sewage. The Rouge and Detroit Rivers
           received most of the CSOs, with 1,296 incidents.

Efforts Have Been Made to Supplement Existing Notification Processes

12First Nation tribes are composed of indigenous people of North America
located in what is now Canada.

 EPA's Spill Prevention and Enforcement Efforts Are Limited and the Coast Guard
         Addresses Spill Prevention as Part of Other Compliance Efforts

EPA's Prevention Program Addresses Only Oil Spills

13The regulation applies to non-transportation-related facilities with a
total above ground (i.e., not completely buried) oil storage capacity of
greater than 1,320 gallons, or total completely buried oil storage
capacity greater than 42,000 gallons. In addition to the storage capacity
criteria, to be regulated, a facility, due to its location, must
reasonably be expected to discharge oil into navigable waters of the
United States or adjoining shorelines, into the waters of the contiguous
zone, or that may affect federal natural resources.

14Alternatively, they may explain in the SPCC plan why such measures are
not practicable.

15Facilities subject to Michigan DEQ's Part 5 Rules are those that meet
the definition of oil storage or on-land facility and receive, process,
manufacture, use, store, or ship oil, salt, or polluting materials above
the specified threshold management quantities (TMQs). The TMQs are: total
above ground capacity of 1,320 gallons or one tank larger than 660 gallons
of oil; five solid tons or 1,000 gallons of salt; and outdoor use or
storage of 440 pounds (or indoor use or storage of 2,200 pounds) of
polluting materials.

EPA Is Uncertain As to Which Facilities Are Regulated by Its Spill Prevention
Program, and Conducts Varying Numbers of Inspections Per Year

16Section 91.1 of the act requires implementing regulations that were not
finalized as of the issuance of this report.

                                      Material             
            Facility      Vessel      transfer       Total 
Year  inspections inspections   inspections inspections 1994 92 157 37 286 
1995           63         172            36         271 
1996           42         181            33         256 
1997           22         166            44         232 
1998           17         151            19         187 
1999           16          89            40         145 
2000           14          94            62         170 
2001           29         122            33         184 
2002           31         152            14         197 
2003           31          99            10         140 
2004          131         100            -a         231 
Total         488       1,483           328       2,299 

EPA Inspections Uncovered Significant Spill Prevention Deficiencies, Whereas the
Coast Guard's Inspections Revealed Minor Issues

17The Environmental SWAT Team is a specialized group that can be rapidly
deployed to address critical environmental issues. It complements ongoing
work that the Ministry of Environment, including its district offices,
pursues, including inspection and enforcement activities. The 2004
inspections in the Sarnia area were the result of concerns raised by
chemical spills that occurred in 2003.

EPA and the Coast Guard Issued a Total of 16 Penalties during the Period
Reviewed

18Coast Guard officials stated that penalties are determined by hearing
officers based upon multiple factors, including the size of the spill, the
product discharged, the environmental or economic sensitivity of the
impacted area, the measures in place by the responsible party to prevent
discharges, the steps taken by the responsible party as a result of the
discharge toward preventing further discharges, as well as the financial
impact of the cleanup operations.

19Bill 133, also known as the Environmental Enforcement Statute Law
Amendment Act, was passed by the Ontario Legislature on June 9, 2005; this
legislation provides the Ministry of Environment with the authority to
issue penalties for spills.

                                  Conclusions

20These funds may also be utilized when the responsible party is known,
but unable to pay for spill response.

21The 2002 Rouge River spill accounts for approximately $7 million of the
total spent in the 11-year time frame.

                      Recommendations for Executive Action

                       Agency Comments and Our Evaluation

Appendix I: Scope and Methodology Appendix I: Scope and Methodology

ADetr Appendix II: Spill Data for the St. Clair-Detroit River Corridor,
1994-2004

                                 EPA Spill Data

                Unknown     Gasoline and                  Waste-              
Year  Oilsa material           fuelsb Chemicalsc Tarsd watere Otherf Total
1994     71       56               20          2     0      0      0   149 
1995     59       22               11          1     2      0      1    96 
1996     51       39                8          2     1      0      0   101 
1997     52       23               14          0     0      0      0    89 
1998     49       49                5          2     1      1      1   108 
1999     41       27               16          1     1      0      1    87 
2000     15       24                5          1     0      0      0    45 
2001     22       17                5          0     0      0      2    46 
2002     19       23                8          0     0      0      0    50 
2003     18       43               12          0     0      0      1    74 
2004     17       43                7          2     0      0      2    71 
Total   414      366              111         11     5      1      8   916 

                             Coast Guard Spill Data

Year  Oils Gasoline and fuels Coal tar Chemical Other Total 
1994     7                  1        0        0     0     8 
1995     2                  0        1        1    1a     5 
1996     2                  0        1        0     0     3 
1997     6                  1        0        1     0     8 
1998     4                  0        0        1     0     5 
1999     4                  2        0        1     0     7 
2000     1                  1        0        0     0     2 
2001     0                 1b        0        1     0     2 
2002     3                  0        0        0     0     3 
2003     2                  0        0        0     0     2 
2004     5                  0        1        0     0     6 
Total   36                  6        3        5     1    51 

                                  Gasoline and                                
Water body                Oils        fuels Coal tar Chemicals Other Total
St. Clair River              3            0        0         0     0     3 
Lake St. Clair               4           5a        0         1     0    10 
Rouge River                 10            0        0         0    1b    11 
Detroit River               18            1        3         4          26 
Not identified in data       1            0        0         0     0     1 
set                                                                  
Totalc                      36            6        3         5     1    51 

                            Michigan DEQ Spill Data

Year  Chemicals Oils Wastewater Gasoline and fuels Tars Other Total 
1996          0    0          0                  0    1     0     0 
1997          0    0          0                  0    0     0     0 
1998          0    0          1                  0    0     1     2 
1999          0    0          0                  0    0     1     1 
2000          1    0          0                  0    0     1     2 
2001          1    1          1                  1    0     0     4 
2002          1    4          0                  0    0     1     6 
2003          0    0          1                  0    0     1     2 
2004          1    0          1                  1    1     0     4 
Total         4    5          4                  2    2     4    21 

Year  Detroit River Rouge River St. Clair River Total 
1996              1           0               0     1 
1997              0           0               0     0 
1998              0           0               2     2 
1999              0           0               1     1 
2000              1           0               1     2 
2001              2           0               2     4 
2002              2           2               1     5 
2003              1           0               1     2 
2004              3           0               1     4 
Total            10           2               9    21 

                             Ontario SAC Spill Data

                                           Gasoline and                       
Year  Chemicalsa Oilsb Wastewaterc            fuelsd Unknowne Otherf Total
1994           8     1          11                 0        3      3    26 
1995           6     3           2                 0        1      2    14 
1996           5     1           8                 2        0      2    18 
1997          10     2           2                 1        0      2    17 
1998           3     1           5                 1        1      0    11 
1999           4     1           0                 3        2      2    12 
2000           4     4           2                 0        2      0    12 
2001           0     0           1                 0        1      1     3 
2002           0     0           4                 1        0      1     6 
2003           0     2           1                 0       10      2    15 
2004          10     0           3                 1        7      2    23 
Total         50    15          39                 9       27     17   157 

Year  Detroit River Lake St. Clair St. Clair River Total 
1994              5              0              21    26 
1995              4              0              10    14 
1996              3              1              14    18 
1997              6              0              11    17 
1998              2              0               9    11 
1999              6              0               6    12 
2000              6              0               6    12 
2001              1              0               2     3 
2002              3              1               2     6 
2003              7              1               7    15 
2004              4              2              17    23 
Total            47              5             105   157 

Appendix III: CSO Data for the St. Clair-Detroit River Corridor, 1999-2004
Appendix III: CSO Data for the St. Clair-Detroit River Corridor, 1999-2004

Table 10: CSOs in the St. Clair-Detroit River Corridor from 1999-2004 by
Water Body

                   Treated                                                      
                        to Diluted Partially                              
Water     Fully   NPDES     raw   treated    Raw       Not          No 
body    treated   limit  sewage    sewage sewage specified information Total
Detroit       0                                                              
River                39     316       154      1        84           0   594
Rouge         0                                                              
River                 0     508        98     17        78           1   702
St.           0                                                              
Clair                                                                  
River                 0      91         0      0        38           0   129
Lake          0                                                              
St.                                                                    
Clair                 0       6        16      9        22           0    53
Total         0      39     921       268     27       222           1 1,478 

Source: GAO analysis of Michigan DEQ data.

CSOs accounted for over 900,000 million gallons of partially treated
sewage discharged into waters of the corridor.

Table 11: Volume of CSOs in the St. Clair-Detroit River Corridor from
1999-2004 (in Millions of Gallons)

                  Volume of diluted Volume of raw Volume of partially         
Water body            raw sewage        sewage      treated sewage   Total
Detroit River              4,741         1,718              26,588  33,047 
Rouge River                3,254            50             885,846 889,149 
St. Clair                                                                  
River                        233             0                   0     233
Lake St. Clair               526           437                 525   1,488 
Total                      8,753         2,205             912,959 923,917 

Source: GAO analysis of Michigan DEQ data.

Appendix IV: NPDES Industrial Effluent Violation Data for the St.
Clair-Detroit River Corridor, 1994-2004 Appendix IV: NPDES Industrial
Effluent Violation Data for the St. Clair-Detroit River Corridor,
1994-2004

The National Pollutant Discharge Elimination System (NPDES) requires
industrial and municipal facilities to obtain permits to discharge
pollutants into U.S. waters. Such permits establish required effluent
limitations or best management practices. The industrial effluent
violation data we obtained from EPA rely upon self-reporting by
industries, and therefore the data have the same data reliability
limitation as spills and CSO data in terms of uncertainty about whether
all events are reported. In addition, volumes are not commonly reported
with effluent discharge violations as toxicity is a greater concern-and
therefore volume data are limited. However, the data provide additional
information on pollutant discharges in the corridor. From 1994 through
2004, there were a total of 2,257 NPDES industrial effluent violations in
the St. Clair River, Lake St. Clair, Detroit River, and Rouge River. Of
these violations, 1,871 (or about 83 percent) of the total had volumes of
greater than 50 gallons (or of an unknown volume). The two largest NPDES
discharge violations, in terms of volume, related to oil and grease-and
these were discharged by the same facility in 1994 only a few months
apart. The most frequently discharged materials were solid pollutants,
pH-altering materials, oil and grease, and materials that had the
potential to alter oxygen availability in the receiving waters. Solid
pollutants include pollutants found in wastewater that were not removed
during the treatment process and can cause toxic conditions or contaminate
sediment.

From 1994 through 2004 the volume of discharged materials was available
for 204 of the 1,871 permit violations. For the remaining 1,667 (or 89
percent) of the violations, the volume was not available.

Table 12: Number of NPDES Industrial Effluent Violations by Volume per
Year

Year  Number of NPDES violations Volume per year if known (in gallons) 
1994                         271                            23,073,813 
1995                         252                                18,157 
1996                         279                                23,707 
1997                         232                                26,911 
1998                         179                                 7,492 
1999                         169                                 9,376 
2000                         102                                 7,934 
2001                         130                                17,769 
2002                         121                                13,133 
2003                          79                                 5,647 
2004                          57                                42,085 
Total                      1,871                            23,246,024 

Source: GAO analysis of EPA data.

Over 50 percent of the materials discharged by industries in violation of
their permits were solid pollutants, oil and grease, zinc, or materials
that alter the pH or oxygen available in the receiving waters into which
they were discharged.

                                                        Zinc,                 
              Solid     Oil and  Carbonaceous 05 day,   total     Other 
Year  pollutants  pH  grease               20c BOD (as ZN) materials Total
1994          37  44      37                    21      30       102   271 
1995          54  39      11                    23      12       113   252 
1996          48  34      23                    18      15       141   279 
1997          46  29       9                    10      19       119   232 
1998          47  31       4                    13       5        79   179 
1999          30  33      16                     9       7        74   169 
2000          30  12       4                     2       0        54   102 
2001          30  25      15                     3       2        55   130 
2002          34   8      15                     3       5        56   121 
2003          11  11       7                    10       7        33    79 
2004          12   9       6                     7       1        22    57 
Total        379 275     147                   119     103       848 1,871 

Table 13: Number of NPDES Industrial Effluent Violations by Discharged
Material and Year

Source: GAO analysis of EPA data.

Over 52 percent of the NPDES violations occurred at 12 facilities, and 1
facility had 176 violations during the 11-year time frame.

Table 14: Number of NPDES Industrial Effluent Violations by Water Body and
Material for the 12 Most Frequently Violated Permits

                                 Materials          Receiving water             Number of 
Facility name    Facility types  discharged         location        Years      violations 
Detroit Metro    Airports,       Phosphorus, pH,    Frank & Poet    1994,             176 
Wayne County     flying fields,  oxygen dissolved,  Drain/Sexton    1995,      
Airport          and service     oil and grease,    Kilter-         1996,      
                                 nitrogen ammonia,  (Detroit River) 1997,      
                                 lagoon freeboard,                  1998,      
                                 carbonaceous                       1999,      
                                 biological oxygen                  2000,      
                                 demand                             2001,      
                                                                    2002,      
                                                                    2003, 2004 
DSC- Trenton     Blast furnace   Zinc, total        Trenton         1994,             131 
plant            and steel works suspended solids,  Channel-Detroit 1995,      
                                 total recoverable  River           1996,      
                                 phenolics, pH, oil                 1997, 1998 
                                 and grease, lead,                             
                                 free cyanide,                                 
                                 cyanide total                                 
Chrysler-Chelsea Commercial      Total suspended,   Mill Creek      1994,             129 
proving grounds  testing         solids, phosphorus (Detroit River) 1995,      
                 laboratory      total removal,                     1996,      
                                 dissolved oxygen,                  1997,      
                                 pH, oil and                        1998,      
                                 grease, nitrogen                   1999,      
                                 ammonia,                           2000,      
                                 carbonaceous                       2001,      
                                 biological oxygen                  2002,      
                                 demand, coliform                   2003, 2004 
                                 fecal material                                
United States    Blast furnace   Zinc, final        Detroit River   1994,             113 
Steel- Ecorse    and steel works toxicity, total                    1995,      
                                 suspended solids,                  1996,1997, 
                                 pH, oil and                        1998,      
                                 grease, copper                     1999,      
                                                                    2000,      
                                                                    2001,      
                                                                    2002,      
                                                                    2003, 2004 
Mobile Home Park Operation of    Total suspended    Huron River     1995,              75 
& Marina         mobile home     solids, phosphorus (Detroit River) 1996,      
                 sites           total percent,                     1997,      
                                 coliform fecal                     1998,      
                                 material, chlorine                 1999,      
                                 total residual, 5                  2000,      
                                 day biological                     2001,      
                                 oxygen demand                      2002,      
                                                                    2003, 2004 
Cargill salt     Chemical and    Total suspended    St. Clair River 1994,              63 
division- St.    chemical        solids, pH,                        1995,      
Clair            preparation     conductivity,                      1996,      
                                 chloride                           1997,      
                                                                    1998,      
                                                                    1999,      
                                                                    2000,      
                                                                    2001,      
                                                                    2002,      
                                                                    2003, 2004 
Selfridge Air    Airport, flying Oil and grease,    Clinton River   1994,              57 
National Guard   fields, and     pH, dissolved      and Lake St.    1995,      
Base             service         oxygen,            Clair           1997,      
                                 carbonaceous                       1999,      
                                 biological oxygen                  2001,      
                                 demand                             2002,      
                                                                    2003, 2004 
Arkema Inc.      Industrial      Triethylamine,     Trenton Channel 1994,              56 
                 inorganic       zinc, total        of the Detroit  1995,      
                 chemicals       suspended solids,  River           1996,      
                                 pH, diethylamine,                  1997,      
                                 carbonaceous                       1998,      
                                 biological oxygen                  1999,      
                                 demand                             2001,      
                                                                    2002,      
                                                                    2003, 2004 
DSC              Nonclassifiable Zinc, total        Frank & Poet    1994,              54 
Ltd.-Gibraltar   establishments  suspended solids,  Drain (Detroit  1995,      
                                 oxygen demand sum  River)          1996,      
                                 product, dissolved                 1997,      
                                 oxygen, oil and                    1998,      
                                 grease, nitrogen                   1999,      
                                 ammonia,                           2000,      
                                 carbonaceous 5-day                 2001,      
                                 biological oxygen                  2002, 2003 
                                 demand, lead, pH                              
Michigan         Steel pipe and  Vanadium, zinc,    Yerkes Drain    1994,              52 
Seamless Tube    tubes           final toxicity,    (Detroit River) 1995,      
LLC                              total suspended                    1996,      
                                 solids, total                      1997,      
                                 phosphorus, pH,                    1998,      
                                 oil and grease,                    1999,      
                                 total cadium                       2000,      
                                                                    2001,      
                                                                    2002,      
                                                                    2003, 2004 
E.B. Eddy Paper  Paper mills     Total suspended    St. Clair &     1994,              43 
Inc.                             solids, pH,        Black Rivers    1995,      
                                 copper, total                      1996,      
                                 residual chlorine,                 1997,      
                                 5 day biological                   1998,      
                                 oxygen demand                      1999,      
                                                                    2000,      
                                                                    2001,      
                                                                    2002,      
                                                                    2003, 2004 
Trelleborgysh    Rubber and      Oil and grease,    Black River via 1994,              46 
Inc.-Sandusky    plastic hoses   total residual     Stone Drain     1995,      
                 and belts       chlorine, toulene, (St. Clair      1996,      
                                 dichloroethylene,  River)          1997,      
                                 trichloroethylene,                 1999,      
                                 pH, toxicity,                      2000,      
                                 dissolved oxygen,                  2001,      
                                 phosphorus total                   2002,      
                                                                    2003, 2004 
Other permit                                                                          876 
violations                                                                     
Total violations                                                                    1,871 

Source: GAO analysis of EPA data.

Appendix V: Spill Profiles for Six Selected Incidents Appendix V: Spill
Profiles for Six Selected Incidents

                   U.S. Spill: 2001 Hazardous Substance Spill

           o  From May 16 at 1:00 p.m. to May 17 at 1:00 p.m., two spills of
           approximately 14-15 million gallons of storm water mixed with
           ethylene glycol and propylene glycol (deicing agents) were
           discharged into a storm sewer leading to the Detroit River.1 The
           responsible party claimed that the release was due to blockage in
           a 10-inch pipe from a holding pond containing the material to the
           sanitary sewer system.

           o  Initially, EPA estimated that 10,000 fish were killed due to
           depletion of dissolved oxygen in the waterway.

           o  On May 18, the NRC received a spill report from an observer who
           saw fish dying. NRC reported the spill to EPA, the Coast Guard,
           and Michigan DEQ shortly after 6:00 p.m. At 6:53 p.m., EPA
           contacted the Michigan Pollution Emergency Alerting System (PEAS)
           hotline. At 8:00 p.m., the PEAS operator contacted Michigan DEQ
           Water Bureau staff. At 9:00 p.m., the EPA on-scene coordinator
           notified the Michigan DEQ and the Coast Guard. When the Michigan
           DEQ spill responder did not arrive on-scene, the PEAS operator
           called the Michigan DEQ District Supervisor at 11:30 p.m. On May
           19 at 12:15 a.m., the Michigan DEQ District Supervisor contacted a
           spill responder, saying that EPA had been on the scene and was
           requesting Michigan DEQ representation. At 8:00 a.m. on May 19,
           DEQ staff arrived at the scene. At 10:30 a.m., two Coast Guard
           responders arrived at the scene.

           o  At 5:00 p.m. on May 20, tanker trucks flushed out an isolated
           section of the affected sewer drain with clean water. A pump was
           installed to pump water to the nearby wastewater treatment
           facility. This lasted until 10:00 a.m. on May 21.

           o  The responsible party notified Michigan DEQ of the May 16 and
           17 discharges on May 22.

           o  The facility that is responsible for these discharges has an
           industrial NPDES permit. Michigan DEQ agreed to accept best
           management practices instead of numeric pollutant limits for the
           summer discharges from this facility. So from May through
           September, the facility's permit had no limitations on
           oxygen-depleting materials. Michigan DEQ's understanding was that
           all discharges containing high amounts of oxygen-depleting
           materials would be directed to the sanitary sewer, for further
           treatment at the wastewater treatment facility.

           Figure 11: 2002 Rouge River Oil Spill

           o  On April 9, at 9:50 a.m., an observer notified the Coast
           Guard's Detroit Marine Safety Office of an oil spill near their
           property. At 10:00 a.m., the Coast Guard received a spill report
           from a nearby bridge operator. At 10:01 a.m., the Coast Guard
           notified the NRC. At approximately 11:00 a.m., the Coast Guard
           arrived on the scene. At 1:00 p.m., the Coast Guard briefed
           Michigan DEQ. At 2:05 p.m., the Coast Guard notified EPA.
           o  On April 10, at 9:20 a.m., the Coast Guard and EPA discussed
           the spill via phone. At 11:00 a.m., EPA and a contractor arrived
           at Dearborn, MI, to begin emergency assessment activities. Also at
           11:00 a.m., the Coast Guard issued a warning message to Canadian
           officials. At 1:44 p.m., the Coast Guard began contacting Michigan
           drinking water facilities. At 7:55 p.m., the U.S. and Canadian
           Coast Guards agreed to invoke the Joint Marine Pollution
           Contingency Plan.

           o  On April 12, another spill occurred in the Rouge River.

           o  According to the National Oceanic and Atmospheric
           Administration's Scientific Support Coordinator's report, the
           spill material was weathered diesel mixed with used synthetic lube
           oil (such as "3 in 1 oil"), mixed with sewage.

           o  A Coast Guard summary from April 26 indicates 66,359 gallons of
           oil were recovered. An EPA spill summary indicates that multiple
           agencies recovered 63,000 gallons of oily water from open water
           and shorelines and more than 771,564 gallons of oil, oily water,
           and oily sludge from the municipal sewers of Dearborn and Detroit.
           EPA's final estimate for total spill volume was 321,000 gallons of
           oil.

           o  The responsible party did not report the spills nor could
           officials identify the party.

           o  A Coast Guard report from July 1 says that $3,989,905 was
           expended for response. Coast Guard officials from the Detroit
           marine safety office noted that there is not a funding mechanism
           for response when the spill originates on one side of the
           international boundary and damage occurs on the other. A Canadian
           report notes that $1,131,550 in Canadian dollars was spent on
           spill response.

           o  EPA's cost of cleanup and response was about $2.5 million. An
           EPA report says that the Coast Guard, EPA, and responding city
           governments spent approximately $10 million on spill response and
           investigation.

           o  According to the Fish and Wildlife Service (FWS), 17 miles of
           U.S. shoreline were affected and 16 kilometers (10 miles) of
           Canadian shoreline were impacted. Two of the islands with
           shorelines that were oiled were part of the newly created Detroit
           River International Wildlife Refuge.
           o  Over 100 birds and reptiles were reported as dead,
           rehabilitated, or observed oiled or otherwise impacted. A FWS
           representative believes this number significantly underestimates
           the numbers impacted. She believes a multiplier should be used,
           and estimated the actual number of birds and reptiles impacted at
           approximately 720 to 900.
           o  Per the Coast Guard documents, the Captain of the Port closed
           the Rouge River temporarily. Six vessel transits were delayed, one
           cancelled, and two diverted to other ports as a result of the
           incident. Two companies expressed concerns about staying open if
           Rouge River closure extended past April 18. Some recreational
           boats and marinas were oiled.
           o  On the Canadian side of the corridor, a drinking water intake
           advisory was in effect on April 11.

           Figure 12: 2004 Rouge River Oil Spill

           o  On August 3, at 7:00 a.m., a bridge operator notified the Coast
           Guard's Detroit marine safety office of an oil sheen in the Rouge
           River. At 8:00 a.m., the Coast Guard dispatched pollution
           investigators to the scene. At 9:07 a.m., the Coast Guard notified
           Michigan DEQ. At 9:20 a.m., the Coast Guard notified the Ontario
           SAC. At 9:27 a.m., the Coast Guard notified EPA. At 9:30 a.m., the
           Coast Guard notified the NRC.

           o  At 10:05 a.m., a Coast Guard contractor started containment
           efforts. The contractor diverted oil to a collection point and
           boomed the north side of the Rouge River. Vacuum trucks and
           skimmers recovered oil on the southern side of the Rouge River. At
           11:30 a.m., the contractor deployed an additional boom at the
           mouth of the Rouge River to collect any oil that escaped.

           o  At 11:40 a.m., the Detroit Water and Sewage Department
           confirmed multiple CSO discharges into the Rouge River, the last
           of which ended at 9:55 a.m. At 2:10 p.m., the Coast Guard's
           District 9 reported that Canadian authorities reported oil and
           debris on their side of the Detroit River in the vicinity of
           Fighting Island, and it was consistent with a CSO release. At 3:24
           p.m., the NRC notified the Coast Guard and Michigan DEQ of spill
           details provided by the Canadian Coast Guard. The Canadian Coast
           Guard indicated that the Joint Marine Pollution Contingency Plan
           was invoked. At 5:56 p.m., the NRC notified the Coast Guard and
           Michigan DEQ of spill details provided by an observer who stated
           that a black slick was seen coming from a freighter which recently
           took on a load of salt. At 6:03 p.m., the NRC notified the Coast
           Guard and Michigan DEQ of spill details provided by a power plant
           regarding a sheen coming into the plant's inlet canal. On August
           4, the NRC notified the Coast Guard and Michigan DEQ of an oil
           sheen reported again by the power plant.

           o  The spill material was an oil and sewage mixture. The Coast
           Guard originally estimated that the spill was 100 gallons, however
           in their database it is listed as a 2,000-gallon spill. The Coast
           Guard's final estimate was 5,000 gallons.

           o  The responsible party did not report the spills, nor could
           officials identify the party.

           o  EPA officials stated that officials from the City of Dearborn,
           the City of Detroit, Michigan DEQ, EPA, and the Coast Guard
           inspected CSO structures to the Rouge River in the area of
           concern, and no evidence of oil accumulation was found in any of
           the sewers.

           o  On August 3, a project cost ceiling was set at $25,000. This
           was raised on August 4 to $75,000. On August 6, the ceiling was
           raised from $110,000 to $195,000.

           o  On August 28, another oil spill occurred in the Rouge River. An
           EPA report indicates that the individual identifying the spill
           stated that CSOs containing oil had been occurring for 15 months
           intermittently.

           o  The City of Detroit was held responsible for the August 3 and
           August 28 spills, and ultimately paid approximately $249,000 for
           response efforts.

           o  On August 14, at approximately 4:45 p.m., 34 gallons of vinyl
           chloride monomer were discharged into the St. Clair River.2 The
           spill lasted for almost 12 hours. On the following day, another
           spill of 5 gallons of this substance was discharged into the
           river. The cause of the spill was a cracked tube in a cooling
           water system heat exchanger. The responsible party did not report
           the spill to the Ontario SAC until August 19, because an
           electrical blackout caused monitoring equipment to be inoperable.

           o  Ontario Ministry of Environment staff implemented procedures to
           warn downstream intakes and take samples. All samples at Canadian
           reservoirs came back negative. In addition, the Ministry of
           Environment ran models to determine potential impacts. Ministry of
           Environment officials did not issue an advisory, but Chatham
           Health Unit did issue a bottled-water advisory for Wallaceburg
           municipal supply consumers.

           o  Models run by the Ministry of Environment showed that vinyl
           chloride levels would be below the drinking water standards (2
           parts per billion).

           o  There are 12 intakes serving Michigan public water systems in
           the St. Clair watershed between Port Huron and Detroit. Michigan
           DEQ scientists reviewed the incident and determined that the
           amount of vinyl chloride lost, based on a spill of 650 lbs., would
           not have resulted in concentrations at Michigan drinking water
           plant intakes exceeding the maximum contaminant level and that no
           human health risks resulted from the event. No sampling of
           Michigan drinking water plant intakes was conducted upon
           notification of the incident because data collected would not have
           been useful due to the rapid flow rate of the river at the time of
           the event.
           o  On February 1, from 3:00 to 4:20 a.m., an estimated 39,626
           gallons of methyl ethyl ketone and methyl isobutyl ketone were
           discharged into the St. Clair River.

           o  At 5:31 a.m., the responsible party reported to the Ontario SAC
           that they had identified a leaking heat exchanger at the lube
           plant, which resulted in contamination of their cooling water. At
           6:40 a.m., SAC staff briefed the Michigan State Police on the
           incident. At 7:22 a.m., the Michigan DEQ's Pollution Emergency
           Communications Coordinator contacted the relevant Michigan DEQ
           staff. Michigan DEQ staff contacted the SAC for more information
           at 7:45 a.m. Michigan DEQ then notified Michigan drinking water
           facilities between 8:00 and 9:00 a.m. After 11:00 a.m., the
           Michigan DEQ made a decision to recommend that drinking water
           facilities shut their intakes. Drinking water facilities in Port
           Huron, Marysville, St. Clair, East China Township, Marine City,
           Algonac, Ira Township, New Baltimore, Mt. Clemens, Grosse Pointe
           Farms, Highland Park, and Wyandotte were advised of the situation
           and all plants except Port Huron, the Detroit plants, and
           Wyandotte were asked to shut down by Michigan DEQ.

           o  The spill caused more than a dozen water plants on either side
           of the river to close their intakes. About 36,000 customers in the
           St. Clair and Macomb County communities of Marysville, St. Clair,
           East China Township, Marine City, Algonac, and Ira Township were
           adversely impacted by the intake closures.

           o  On May 23, at 4:10 a.m. and 6:00 a.m., an unknown number of
           gallons of oily water were discharged into the St. Clair River.
           Heavy rains caused three oil separators to overflow.

           o  At 6:05 a.m., the responsible party reported the spills to the
           Ontario SAC. The responsible party began sampling and told Ontario
           officials that there were no visible signs of oil or contaminants.
           At 7:40 a.m., the Ontario SAC notified Michigan officials through
           the PEAS hotline. From 8:30 to 9:30 a.m., a Michigan DEQ official
           notified Michigan drinking water facilities.

           The Ohio River Valley Sanitation Commission (ORSANCO) was
           established in 1948 in order to control and abate pollution in the
           Ohio River Basin. ORSANCO is an interstate commission representing
           eight states and the federal government. Member states include
           Illinois, Indiana, Kentucky, New York, Ohio, Pennsylvania,
           Virginia, and West Virginia. ORSANCO has programs to improve water
           quality in the Ohio River and its tributaries. Their tasks include
           setting wastewater discharge standards, performing biological
           assessments, monitoring the chemical and physical properties of
           the waterways, and conducting special surveys and studies. In
           addition, ORSANCO coordinates emergency response activities for
           spills or accidental discharges to the river and coordinates
           public participation in programs.

           In 1977, an unreported discharge of hazardous chemicals
           contaminated drinking water facilities in the corridor. Due to the
           lack of a coordinated monitoring system, misinformation was
           distributed to the public, causing concern for the safety of the
           drinking water. This incident demonstrated the vulnerability of
           the Ohio River water intakes to spills and led to the development
           of the Organics Detection System. ORSANCO, in conjunction with
           drinking water utilities, identified strategic locations along the
           river where monitoring for chemicals would be most beneficial and
           protective of drinking water intakes. ORSANCO suggested that water
           facilities located at strategic points along the river could
           perform routine monitoring for oil and hazardous chemical
           discharges. ORSANCO proposed that they serve as technical
           coordinator and information clearinghouse, providing statewide
           communications in the event of a spill.

           Currently, ORSANCO maintains an inventory of water intakes,
           wastewater discharges, and material transfers on the Ohio River.
           Also, a time-of-travel model is used to estimate the arrival time
           of contaminant discharges during spill events. The results of the
           model have been used to identify the locations of the Organics
           Detection System. The Organics Detection System was established in
           1978 and participants include 11 water utilities, one chemical
           manufacturer, and one power generating facility. Data from each
           facility is to be downloaded for review and evaluation on a weekly
           basis. Each instrument can detect and quantify twenty-two organic
           compounds. The list of compounds represents the organic chemicals
           of greatest concern to water utilities and most likely to be
           detected based on an inventory of chemicals stored, transported,
           and manufactured along the Ohio River.

1Volume estimate provided by the responsible party.

                           U.S. Spill: 2002 Oil Spill

                           U.S. Spill: 2004 Oil Spill

                 Canadian Spill: 2003 Hazardous Substance Spill

2Volume estimate was provided by the responsible party. Vinyl chloride is
a potential human carcinogen when inhaled, and it has been shown to cause
liver cancer in laboratory animals exposed orally or by inhalation.

                 Canadian Spill: 2004 Hazardous Substance Spill

                         Canadian Spill: 2004 Oil Spill

Appendix VI: ORSANCO Spill Detection and Notification System Appendix VI:
ORSANCO Spill Detection and Notification System

Figure 13: ORSANCO Organics Detection System Monitoring Sites

Facility operators are required to notify ORSANCO when detection of a
compound over a specified threshold is observed or when an unidentified
compound is detected. When this occurs, plant operating personnel are
notified of the contaminant so treatment techniques to remove the compound
can be implemented. ORSANCO notifies downstream water utilities and state
and federal water quality and emergency response agencies, including the
NRC.

Figure 14: Water Quality Monitoring Equipment Used at ORSANCO Sites

Appendix VII: Comments from the Department of Homeland Security Appendix
VII: Comments from the Department of Homeland Security

Appendix VIII: A Appendix VIII: GAO Contact and Staff Acknowledgments

                                  GAO Contact

John B. Stephenson (202) 512-3841

                             Staff Acknowledgments

In addition to the individual named above, Kevin Bray, John Delicath,
Michele Fejfar, Jill Roth Edelson, Katheryn Summers Hubbell, Jamie
Meuwissen, and John Wanska made key contributions to this report.

(360600)

GAO's Mission

The Government Accountability Office, the audit, evaluation and
investigative arm of Congress, exists to support Congress in meeting its
constitutional responsibilities and to help improve the performance and
accountability of the federal government for the American people. GAO
examines the use of public funds; evaluates federal programs and policies;
and provides analyses, recommendations, and other assistance to help
Congress make informed oversight, policy, and funding decisions. GAO's
commitment to good government is reflected in its core values of
accountability, integrity, and reliability.

Obtaining Copies of GAO Reports and Testimony

The fastest and easiest way to obtain copies of GAO documents at no cost
is through GAO's Web site ( www.gao.gov ). Each weekday, GAO posts newly
released reports, testimony, and correspondence on its Web site. To have
GAO e-mail you a list of newly posted products every afternoon, go to
www.gao.gov and select "Subscribe to Updates."

Order by Mail or Phone

The first copy of each printed report is free. Additional copies are $2
each. A check or money order should be made out to the Superintendent of
Documents. GAO also accepts VISA and Mastercard. Orders for 100 or more
copies mailed to a single address are discounted 25 percent. Orders should
be sent to:

U.S. Government Accountability Office 441 G Street NW, Room LM Washington,
D.C. 20548

To order by Phone: Voice: (202) 512-6000 TDD: (202) 512-2537 Fax: (202)
512-6061

To Report Fraud, Waste, and Abuse in Federal Programs

Contact:

Web site: www.gao.gov/fraudnet/fraudnet.htm E-mail: [email protected]
Automated answering system: (800) 424-5454 or (202) 512-7470

Congressional Relations

Gloria Jarmon, Managing Director, [email protected] (202) 512-4400 U.S.
Government Accountability Office, 441 G Street NW, Room 7125 Washington,
D.C. 20548

Public Affairs

Paul Anderson, Managing Director, [email protected] (202) 512-4800 U.S.
Government Accountability Office, 441 G Street NW, Room 7149 Washington,
D.C. 20548

www.gao.gov/cgi-bin/getrpt?GAO-06-639.

To view the full product, including the scope

and methodology, click on the link above.

For more information, contact John B. Stephenson at (202) 512-3841 or
[email protected].

Highlights of GAO-06-639, a report to congressional requesters

June 2006

CLEAN WATER

Better Information and Targeted Prevention Efforts Could Enhance Spill
Management in the St. Clair-Detroit River Corridor

Spills of oil and hazardous substances in the St. Clair-Detroit River
corridor have degraded this border area between the United States and
Canada and are a potential threat to local drinking water supplies. Within
the United States such spills are reported to the National Response Center
(NRC), and in Canada to the Ontario Spills Action Centre. This report
discusses (1) how many oil and hazardous substance spills greater than 50
gallons (or of an unknown volume) were reported in the corridor from 1994
to 2004, and how accurately reported spills reflect the extent of actual
spills; (2) what processes are used to notify parties of spills, and if
they contain explicit requirements for reporting times and spill
magnitude; and (3) the extent of Environmental Protection Agency (EPA) and
the Coast Guard's spill prevention efforts and enforcement activities in
the corridor from 1994 to 2004.

What GAO Recommends

GAO recommends that EPA Region 5 and the Coast Guard's District 9 update
spill information and determine whether existing spill notification
processes can be improved. GAO also recommends that EPA consider gathering
information on which facilities are regulated under its spill prevention
program and develop goals for its spill prevention inspections. The
Department of Homeland Security agreed with our findings and conclusions
overall and EPA provided technical comments only.

The NRC received 991 spill reports and the Ontario Spills Action Centre
received 157 reports of spills in the corridor from 1994 through 2004, but
these reports do not accurately portray the actual number or volume of
spills. Many spills go unreported by responsible parties because they do
not understand or fail to comply with reporting requirements. Further,
multiple reports for the same spill are often recorded by NRC and provided
to EPA and the Coast Guard for investigation. EPA does not remove all
duplicate spill reports or update its data after investigating spills.
Coast Guard officials update their spill data after investigations but
they are unable to update spill volume estimates due to automated system
limitations. GAO also found that, according to agency data sets, other
events-combined sewer overflows (CSOs) and industrial permit
violations-occurred more frequently than spills in the corridor. While
data on industrial permit violations and CSOs might be subject to the same
limitations as the spill data because the data are self reported and
facilities may not report all of these events, spills may be particularly
subject to underreporting because they are not part of a structured
program as CSOs and industrial permit violations are.

There are multiple parties involved in spill notification in the corridor
and agreements outlining U.S.-Canadian notification processes are not
explicit about reporting times or the magnitude of spills that warrant
notification. The coast guards of each country have agreed to notify one
another of spills primarily when a joint response may be needed. Another
agreement between Michigan and Ontario officials calls for notifying each
other of spills that may have a joint impact. We reviewed six selected
spill incidents that illustrate the various ways that notification can
occur. The drinking water facility operators we contacted on the U.S. side
of the corridor had differing perspectives on current notification
processes, and the majority expressed concern that their facilities could
be contaminated by spills if they are not notified in a timely manner.
Finally, efforts have been made to develop informal notification processes
between individual industries or trade associations and drinking water
facilities.

EPA's spill prevention program is limited and the Coast Guard addresses
spill prevention as part of other compliance efforts. EPA's prevention
program addresses only oil spills. Further, EPA is uncertain of which
specific facilities are subject to regulation under its spill prevention
program, and conducts varying numbers of inspections per year. EPA
inspections uncovered significant spill prevention deficiencies, whereas
the Coast Guard's inspections revealed minor issues. The agencies issued a
total of 16 penalties for spills and program noncompliance during the
period we reviewed.
*** End of document. ***