v> U.S. Environmental Protection Agency EVALUATION OF THE RESPONSE TO THE MAJOR OIL SPILL AT THE ASHLAND TERMINAL, FLOREFFE, PENNSYLVANIA BY THE INCIDENT-SPECIFIC REGIONAL RESPONSE TEAM ------- EVALUATION OF THE RESPONSE TO THE MAJOR OIL SPILL AT THE ASHLAND TERMINAL, FLOREFFE, PA BY THE INCIDENT-SPECIFIC REGIONAL RESPONSE TEAM ------- EVALUATION OF THE RESPONSE TO THE MAJOR OIL SPILL AT THE ASHLAND TERMINAL, FLOREFFE, PA BY THE INCIDENT-SPECIFIC REGIONAL RESPONSE TEAM TABLE OF CONTENTS Page Introduction i Executive Summary ii Section I: Description of the Incident 1 Section II: Description of the Response 3 Section III: Findings and Recommendations 10 Section IV: Conclusion 18 List of Appendices Water User Experience Transcript of the NRC Report of the Ashland Oil Spill, Floreffe, Allegheny County, PA Incident-Specific Regional Response Team River Monitoring Work Group Glossary of Abbreviations Assessment of Monitoring Elements Appendix A: Appendix B: Appendix C: Appendix D: Appendix E: Appendix F: ------- INTRODUCTION ------- INTRODUCTION A four-million gallon steel storage tank collapsed at the Ashland Oil Company terminal in Floreffe, Pennsylvania, Saturday Evening, January 2, 1988, creating one of the largest inland oil spills in the United States. A wave of 3.9 million gallons of diesel fuel crashed over the tank's containment dike and surged in all directions. Initial estimates were that over 1,000,000 gallons escaped into the Monongahela River through storm sewers located on an adjacent property. These estimates were later revised to approximately 750,000. This report provides the Incident-Specific Regional Response Team's (RRT's) evaluation of the response to the Ashland Oil Spill. Five work groups were established to assess the response and their draft submissions are the basis of this report. The five work groups assessed the initial response, water supply, river monitoring, river cleanup, and communications. This report provides relevant background information pertaining to the incident and the response, as well as the RRT's findings and recommendations for improving future responses. Organization of the Report Section I provides background information relating to the January 2 release, as well as the natural and technological factors affecting the emergency response. Section II provides a description of the initial on-site response, river monitoring, and river cleanup activities. Section III presents the RRT's findings regarding the emergency response activities and provides specific recommendations to improve the efficiency and effectiveness of future responses. Section IV outlines the RRT's overall conclusions regarding the response to the Ashland Oil Spill. i ------- EXECUTIVE SUMMARY ------- EXECUTIVE SUMMARY Description of the Incident A total of 3,881,841 gallons of No. 2 diesel fuel were released from a collapsing oil storage tank at the Ashland Oil Company terminal in Floreffe, Pennsylvania on Saturday, January 2, 1988 at 5:10 p.m. The catastrophic release created a wavelike surge of oil that passed over the banks of the facility's containment berms and into a nearby storm drain. Current estimates are that approximately 750,000 gallons of the diesel fuel entered the Monongahela River. Several factors made response efforts difficult. First, due to the force of the release, most of the oil entered the river within the first two hours of the tank rupture, leaving little time to obtain equipment and personnel to stop the flow. Second, the rupture occurred after dark, preventing first responders from conducting a thorough assessment of the extent of the spill. This problem was aggravated by the disconnection of power and communication lines at the site and the evacuation of 250 nearby residents in response to a report of a gasoline leak from an adjacent tank. Third, extremely cold air and water temperatures affected all aspects of the response, including monitoring and cleanup activities. Description of the Response Ashland Oil Company employees provided prompt notification of the tank rupture and spill to the National Response Center and to local emergency response agencies. Local authorities were the first on-scene responders. Early response efforts by these authorities and Ashland's cleanup contractor were directed toward preventing the spilled diesel fuel from entering the Monongahela River by blocking off storm drains and by creating temporary containment dikes. The U.S. Coast Guard (USCG) acted as the first federal official on scene and was joined at daybreak by the U.S. Environmental Protection Agency's (EPA's) On-Scene Coordinator (OSC). The OSC advised Ashland that EPA would direct and monitor all phases of the cleanup operation. Ashland agreed to EPA's direct control and also agreed to assume full responsibility for all cleanup costs incurred. The Incident-Specific Regional Response Team (RRT) was formally activated on Monday, January 4. Prior to that time, however, many of the RRT member agencies were already actively involved in the initial response. RRT members acted as conduits of information to and from their respsective agencies and made recommendations for appropriate response actions. ii ------- State agencies primarily directed their efforts toward river water quality monitoring and concerns regarding the provision of safe drinking water supplies. To those ends, three types of river monitoring took place: Monitoring at water utility intakes to protect water supplies, monitoring spill effects on fish and wildlife, and monitoring of the river to define the spill mass and movement. The USCG coordinated and oversaw the recovery of oil in both the Monongahela and Ohio Rivers. The river cleanup operations spanned 38 miles and were severely hampered by near arctic weather conditions. In addition, as the oil passed through the series of locks and dams along the rivers, the oil became dispersed throughout the water column limiting boom containment to the oil remaining on the water's surface. Summary of Findings and Recommendations Coordination During the initial eighteen (18) hours of the response, local, state, and federal agencies worked independently to minimize the dangers and damage resulting from the spill, and no single agency assumed an overall coordinating role. The RRT found that with one exception, this limited response coordination did not concern the agencies at the time since each had numerous individual functions to perform. After the initial 18 hours, coordination improved, but was still hampered in part by the widely separate areas where response activities needed to occur. The RRT concluded that either a single, large command post or better liaison among multiple response sites would have enhanced response coordination. The RRT recommends in the future that earlier coordination and an assignment of responsibilities by the OSC occur. The Region III Contingency Plan states that "the RRT shall be activated automatically in the event of a major or potentially major discharge or release", but the RRT was not activated until Monday, January 4. The RRT concludes that it could have provided valuable assistance to the responding agencies had the RRT been involved in the first 24 hours following the spill. The RRT recommends that the team be activated as soon as possible after a large spill occurs, and that a decision should be made as to whether or not it would be necessary for RRT members to be brought together on site. The RRT further recommends the designation of an "RRT Coordinator" to assist the OSC by facilitating communications among responding agencies. iii ------- The RRT finds that in the initial days of the spill, the coordination and communications of river monitoring data suffered because no lead agency was assigned to oversee these activities. The RRT recommends for future responses the early establish- ment of a single coordinating agency to focus the coordination and communication of monitoring data so that relevant data is efficiently collected, and to ensure a standardization of analysis. Communications Communications problems were encountered during the Ashland Oil Spill due to insufficient communications equipment at the command post to support the large number of responding agencies. The RRT recommends immediate installation of an adequate number of phone lines to support these agencies at command posts in addition to the use of cellular phones. The RRT further recommends that an efficient means of communications between the command post and RRT members be developed by the RRT Communications Work Group. Resources The lack of immediately available containment and monitoring equipment hindered the response. Delays were caused by the need to locate and transport essential equipment. The RRT recommends the preparation of inventories of locally available equipment to assist emergency responders in quickly locating necessary equipment. The Ashland Oil Spill could have been far more devastating had public water supplies been contaminated or water shortages more severe. The RRT recommends that emergency planning agencies and water suppliers work toward improving the availability of contingency water supplies. iv ------- SECTION I DESCRIPTION OF THE INCIDENT ------- SECTION I DESCRIPTION OF THE INCIDENT On Saturday, January 2, 1988 at approximately 5:10 p.m., 3,881,841 gallons of No. 2 diesel fuel were released from a collapsing oil storage tank at the Ashland Oil Company terminal in Floreffe, Pennsylvania. The Ashland terminal is a small (35- tank) oil and tar tank farm located between State Highway 837 and the Duquesne Power and Light Company, approximately 200 yards from the Monongahela River. The catastrophic release created a wavelike surge of oil that passed over the banks of the facility's containment berms and into a nearby storm drain. Initial estimates were that 1,000,000 gallons of the diesel fuel entered the Monongahela River via a storm sewer running beneath the Duquesne property. These estimates were later revised to approximately 750,000 gallons. Most of the 750,000 gallons of oil that entered the river did so within the first two hours following the rupture of the tank. Even under the most perfect conditions, it would have been difficult to obtain the personnel and equipment necessary to block off the drains and the outfall to the river in such a short time. Several factors generally outside the control of federal, state and local personnel contributed to making conditions at the site less than ideal for responding to a major oil spill. The rupture of the storage tank occurred at dusk. Darkness prevented a thorough assessment at that time of the extent of the spill, and that in turn delayed initial response actions. In response to reports of a gasoline leak, all electrical and telecommunication lines were disconnected at the Ashland terminal to diminish the threat of fire and explosion. Approximately 250 nearby residents were evacuated. The resulting confusion made site access difficult and impeded the efforts of first responders to effectively contain the spill. The Monongahela River current was initially estimated to be moving at 2 mph; however, the actual rate was 1.1 mph, and the flow rate continued to decrease as temperatures fell in the days that followed. Predictions of plume movement were difficult to make due to the changing river flow rate. The spill occurred in the Lock and Dam No. 3 pool at river mile 25. The plume quickly reached Lock and Dam No. 3 which is a little over one mile downstream at river mile 23.8. Passing over this dam and the series of navigational locks and dams on the Monongahela and Ohio Rivers, the oil was emulsified and mixed until it was dispersed throughout the water column. Once the oil was emulsified, it largely escaped containment by booms. Cleanup crews could find few suitable locations for oil collection downstream from the spill site, and river access was impossible at many points. 1 ------- The Pittsburgh area experienced extremely cold weather during the first weeks of 1988. Very low air and water temperatures affected all aspects of response, monitoring, and cleanup activities. The frigid conditions on the river increased the risk of hypothermia for cleanup crews which led to the decision by the OSC to demobilize all personnel from the river on the fourth day after the spill. For several days, the ice cover on the Monongahela and Ohio Rivers was between 50 and 90 percent. The ice, which contained a percentage of the oil, interfered with the placement of booms and sorbent materials and complicated estimations of the movement of the plume based on river velocity. Cold temperatures may have stabilized the oil emulsion in the river. The emulsion in turn, contributed to the difficulty of oil recovery operations and increased the threat of contamination to subsurface water utility intakes. 2 ------- SECTION II DESCRIPTION OF THE RESPONSE ------- SECTION II DESCRIPTION OF THE RESPONSE Notification of Government Agencies Ashland employees provided prompt and thorough notification of the tank rupture and spill to the National Response Center (NRC) and to local emergency response agencies. In turn, the NRC notified the U.S. Coast Guard (USCG) of the spill and at 1810 hours, the USCG notified U.S. EPA Region III. Initial evaluation of the spill took place in darkness and as a result, early reports of the spill greatly underestimated its magnitude. Although all agencies received timely notification, important information about the spill was not known as the agencies prepared their initial response. Initial On-site Response Local authorities were the first on-scene responders and immediately rerouted traffic away from the site and instituted site safety measures. Local mutual aid agreements were exercised and a temporary command post was established at the Floreffe Fire Hall. Early efforts by local response agencies and Ashland's cleanup contractor, O.H. Materials, Inc., were directed toward preventing the spilled diesel fuel from entering the Monongahela River by blocking off storm drains and by creating temporary containment dikes. These efforts were hampered by darkness and extreme cold. The response was further hampered because communication and power lines at the Ashland terminal were shut off as a precautionary measure at the time of the spill. Despite the outstanding efforts of local responders, the plugging, patching, booming, and building of dikes had minimal impact on the flow of oil; there was simply too much moving too quickly. The USCG acted as the first federal official on scene and exercised control of river traffic and mobilized the National Strike Force. After inspecting the site, the USCG provided information to EPA. EPA advised the Coast Guard that the position of Federal On-Scene Coordinator (OSC) would be assumed by EPA and that the OSC would arrive on site at first light. The EPA Region III Technical Assistance Team (TAT) was immediately dispatched to the site. Oversight of Ashland's Response Efforts The USCG monitored Ashland's efforts throughout the initial response phase and directed the river cleanup by providing advice to Ashland. The Coast Guard made the initial determination that Ashland's efforts were proper and thorough. 3 ------- At 0740 hours on January 3, 1988, the EPA OSC arrived on site and delivered a verbal "Notice of Federal Interest to Suspected Discharger" to Ashland officials. The OSC advised Ashland that EPA would direct and monitor all phases of cleanup operations. Ashland agreed to EPA's direct control and also agreed to assume full responsibility for all cleanup costs incurred. The Incident-Specific Regional Response Team (RRT) was formally activated Monday, January 4, 1988, although many of the RRT member agencies were already actively involved in the initial incident response. The following agencies participated in the RRT activations during the incident: U.S. Coast Guard, Second District, Fifth District, MSO Pittsburgh, National Strike Force LANTAREA Strike Team U.S. Environmental Protection Agency (Regions III, IV, V) U.S. Army Corps of Engineers (Pittsburgh, Cincinnati, Huntington, Louisville) U.S. Department of Interior (Philadelphia, Chicago) National Oceanic and Atmospheric Administration (Rockville, Seattle) Occupational Safety and Health Administration Federal Emergency Management Agency Pennsylvania Department of Environmental Resources Pennsylvania Emergency Management Agency West Virginia Department of Natural Resources West Virginia Department of Health Ohio Environmental Protection Agency Ohio River Valley Water Sanitation Commission Kentucky Department for Environmental Protection The RRT was activated to provide advice and guidance to the OSC. RRT members acted as conduits of information to and from their respective agencies. During RRT teleconferences, the political, programmatic, and statutory implications of the actions of the responding agencies were discussed, and recommendations for appropriate response actions were made. 4 ------- Delegation of Tasks During the initial hours of the response, none of the agencies on site assumed the role of "lead response agency." Once on scene the following morning, the OSC assumed the lead role. Specific tasks and responsibilities were assigned by the OSC to the agency best qualified to perform them. State authorities immediately directed their efforts toward concerns over water quality, which by noon on January 3, 1988, became a separate, significant response phase managed by State and county authorities. In particular, the Pennsylvania Depart- ment of Environmental Resources (PADER) and the Pennsylvania Fish Commission (PFC) began assessing the impact on water quality and downstream water intakes. The USCG, with the recognition that Ashland and local authorities were controlling the site response, focused immediately on cleanup and recovery of oil in the river. River Monitoring Three types of monitoring took place during the event: monitoring of the river to define the spill mass and track its movements, monitoring the effects on fish and wildlife, and monitoring at intakes to protect water supplies. Flow and velocity forecasts by the National Weather Service were initially utilized to predict progress of the oil plume. In addition, the U.S. Army Corps of Engineers (USACOE) provided velocity information on a daily basis. Attempts at tracking the plume were initiated on January 4, 1988 and included the culmination of overflight observations and taste and odor reports from treatment plant operators and lock and dam workers. Flow and velocity data and weather forecasts were obtained from appropriate agencies. On January 5, PADER and the PFC began sampling on the Monongahela and Ohio Rivers at three depths in the water column and sediment on the river banks. Analyses performed were Oil and Grease (O&G) and Total Organic Carbon (TOC). PADER, in cooperation with the Allegheny County Health Department, also initiated sampling of surface and groundwater intakes along the rivers in Pennsylvania, with analyses for volatile organics, TOC, and fuel oil performed by the PADER laboratory in Harrisburg. A group at the Ashland command post consisting of represent- atives of U.S. EPA, the National Oceanic and Atmospheric Administration (NOAA), PADER, and the Ohio River Valley Water Sanitation Commission (ORSANCO) set up a program of sampling river water at the first eight water supply intakes downstream of the spill site. The sampling was initially set up as three samples per day at each intake, but had to be cut back to two per 5 ------- day due to logistical considerations. Analyses were first performed at a local contract laboratory (NUS) until a field laboratory could be set up by the cleanup contractor, O.H. Materials, Inc. Analyses included volatile organics, base neutrals, and No. 2 fuel oil. On January 6, personnel from the U.S. EPA, West Virginia Department of Natural Resources (WVDNR), and ORSANCO initiated an effort to track the spill from tow boats. One boat equipped with a fluorometer moved upstream from Wheeling, while a second boat equipped with a TOC analyzer moved downstream from Wheeling. After the first day, it became apparent that the fluorometer provided useful results for characterizing the plume, while the TOC analyzer did not. Thereafter, the effort continued utilizing one tow boat and a fluorometer with personnel from Ohio EPA, WVDNR, and ORSANCO. Samples were collected at the point where the fluorometer indicated the leading edge and the peak concentration of the spill and were shipped to the WVDNR's Guthrie Laboratory for analysis of volatile organics, base neutrals, and fuel oil. On January 6, 7 and 8, the USACOE, Pittsburgh collected samples at several locations including the New Cumberland, Montgomery, Emsworth, Elizabeth and Braddock Dams. Ohio EPA set up a monitoring system at eight sites on the Ohio River and commenced sampling on January 7. Analyses included TOC, O&G, and organics. On January 13, personnel from the USACOE, Huntington District took over the downstream tracking of the spill. They continued to perform the tracking, utilizing their own boat and a flow-through fluorometer, through January 23 when the spill left the limits of their jurisdiction (Meldahl Dam, river mile 436) . The effort was then taken over by the USACOE's Louisville District, which followed the spill until February 2, at which time the spill was no longer detectable by the boat-mounted fluorometer. From January 19 through 23, a tow boat was again employed to assist in the tracking and to provide samples requested by down- stream water users. Five-gallon samples were collected at the indicated peak for use by the utilities performing treatability studies. On January 22, the spill reached Kentucky. Monitoring by the Kentucky Division of Water consisted primarily of fluoro- meters at water intakes. Fluorometers were successfully utilized at Maysville and Louisville. Fluorometers were moved downstream to provide monitoring at water intakes in Evansville, Indiana and Cairo, Illinois. By that time, the spill had dispersed to the point where results were inconclusive. Monitoring at Cairo, just above the mouth of the Ohio River, was concluded on February 12. 6 ------- In general, cold weather hampered the river monitoring efforts. Poor navigation conditions forced the monitoring crews to use slow moving tug boats which slowed the provision of monitoring information to water utilities. Monitoring Effects on Wildlife Dozens of individuals, under the guidance and direction of the Audubon Society and the Pennsylvania Game Commission, worked for days following the spill retrieving and cleaning oil-soaked waterfowl from along the rivers. Efforts to save waterfowl were hampered by low temperatures and by ice on the rivers which kept rescue workers on the shore. Although many birds were saved, estimates of waterfowl mortality range from 2,000 to 4,000 ducks, loons, cormorants, and Canada geese, among others. Biologists from the WVDNR conducted shoreline counts along 120 miles of the Ohio River to determine the number of fish killed. In the week following the spill, several censuses of dead and stressed fish were taken in the dam pools along the river. WVDNR aquatic toxicologists designed oil-impact studies on species of mussels by taking censuses and samples from well- established mussel beds on the Ohio before and after the arrival of the oil slug. In separate studies, mussels and catfish were placed in cages in the river downstream from the spill and were collected after the plume had passed. The organisms will be analyzed to determine the adverse effects of the diesel fuel. Effects on Water Supplies Appendix A outlines by individual water supplier the impacts and actions taken with respect to water supply facilities. In general, however, the provision of information by emergency response agencies enabled downstream water suppliers to implement treatment procedures and increase storage volume before the spill affected their intakes. Public confidence in the water purveyors and in the government agencies responding to the spill was maintained because the quality of the water supplied never deteriorated. Coordination of raw and finished water quality testing and reporting was critical because of the need to use that information in making decisions with respect to water plant operations. Use of private laboratories for quick turnaround time of sample results was vital for determining the effective- ness of carbon treatment and for verifying the quality of drinking water during the incident. 7 ------- Where needed, emergency water stations were set up quickly and refill operations were well organized in spite of cold temperatures which necessitated the staging of water tankers indoors. Special plumbing was rapidly installed so that multiple faucets were available at each station. Dairies and breweries provided bottled water and were given guidelines for protection of the quality of the bottled water. The Commonwealth of Pennsylvania issued a Water Conservation Order for three counties during the incident. This order and the cooperation of the public allowed three of the four affected water suppliers in Pennsylvania to avoid water loss to their customers. River Cleanup It is now known that virtually all the oil entered the river via a single route; a 24" drainage pipe on the adjacent Duquesne Power Company property that discharged into the river at Duquesne's cooling water discharge. Once this route was identified, the fire department installed an underflow dam in front of the drain opening to prevent the oil from flowing into the river. However, due to delays caused by darkness, loss of power and communications, an an evacuation of the area, the dam was not installed until 2100 hours on January 2, 1988. By that time, most of the oil had already reached the river. This first containment was replaced by an inflatable plug that failed early the next morning. At that time, an additional 50,000 gallons of oil were released into the river. The oil spill from the Ashland facility quickly reached the first in a series of locks and dams along the Monongahela and Ohio Rivers. As the oil passed through the locks and dams, it apparently became emulsified and dispersed throughout the water column. As a result, the containment booms were only able to retain a percentage of oil that was not emulsified. Traffic in the Monongahela River was prohibited for 58 hours following the Ashland spill and was restricted for seven days, enabling cleanup crews to move about freely in the river. Daily evaluation of the restrictions on river traffic were made by the Marine Safety Office (MSO) in Pittsburgh. Efforts to contain and recover oil at this enormous spill included the utilization of over 150 people, eleven vacuum trucks, three cranes, and 20,000 feet of river boom. The very extreme weather conditions prevented extensive use of recovery methods such as the application of sorbent materials because ice cover prevented the sorbent materials from contacting the oil. In addition, the ice cover and high risk of injury for the work crews prevented the placing of booms in some otherwise strategic locations. Recovery efforts were centered around diversion and removal of the oil from natural or man-made pockets. 8 ------- River cleanup operations spanned 38 river miles, and were conducted at times in extremely cold conditions. The USCG provided personnel from both MSO Pittsburgh and National Strike Force LANTAREA Strike Team to oversee cleanup operations by contractors hired by Ashland. Cleanup monitoring consisted of visits to the cleanup sites by survey teams, who reported cleanup progress to the On-Scene Coordinator at the command post. Providers of alternative oil recovery technology from all over the world contacted the EPA during the emergency. The following alternative technologies were proposed: biodegra- dation, viscoelastomers, degreasers, demulsifiers, solvent extractors, gelatinizing agents, water purifiers, boat-mounted oil/water separators, robots, bird feathers, and placement of pantyhose in the spillway. All individuals and companies who contacted EPA concerning the use of their products at the spill were referred to the Ashland Oil Company. At the request of Ashland, a field test of the effectiveness of a viscoelastomeric product was performed on the Ohio River during the cleanup. The product neither promoted nor inhibited the rate of oil recovery from the river. 9 ------- SECTION III FINDINGS AND RECOMMENDATIONS ------- SECTION III FINDINGS AND RECOMMENDATIONS The findings and recommendations of the RRT can be generally grouped into three major topics which are critical elements in any emergency response operation: Coordination, Communications and Resources. COORDINATION The initial 18 hours of response to the emergency can be characterized as a loosely organized, but relatively effective attempt, to minimize the dangers and damage from the spill. As local, county, state, and federal agencies responded, each exercised its own responsibilities independently as liaison was established at the scene. At no time in the initial hours did any agency deem it necessary to "take charge" of the entire response effort and only the City of Pittsburgh HAZMAT team expressed some concern about being able to identify the individual in charge. By 8:00 a.m. on January 3, 1988, all responding agencies were on scene and in liaison with the others. A response organization then developed which, while not without some temporary shortcomings, served as an effective structure throughout the response period. This response organiza- tion included the assignment by the OSC of specific areas of responsibility, such as site cleanup, water monitoring, oil recovery, and assistance to water suppliers, to appropriate agencies. Command Post During the Ashland response, there appeared to be a constant tradeoff between the need to have all response personnel in a single command center for better coordination, and the need to limit the number of people present in the command post to minimize the confusion. A single, very large, command post would have enhanced response effectiveness; however, no site for such a center could be located in the first twelve hours, and the locations of the multiple response activities (spill site cleanup, river oil recovery, water supply protection) were widely separated. In retrospect, perhaps an effective future arrange- ment would be to establish a central command post staffed by personnel responsible for each aspect of the response and separ- rate operations centers for each aspect. 10 ------- Regional Response Team The Region III Regional Contingency Plan states: "The RRT shall be activated automatically in the event of a major or potential major discharge or release," but the RRT was not activated until Monday, January 4, 1988. The RRT could have provided assistance to the responding agencies sooner, had it become involved in the response during the first 24 hours following the spill. It is recommended that the RRT become active as soon as possible after future large spills occur. All standing RRT members should be informed of the activation. At the time of RRT activation, a decision should be made about whether or not the effectiveness of the spill response will be increased should RRT members be brought together on site. The RRT should consider designating an "RRT Coordinator" to facilitate communications among responding agencies, including RRT members. During an emergency, such an individual could be in charge of reporting back to the OSC all of the developments at Emergency Operations Centers (EOCs), other than the command post, such as those established during emergencies by the Red Cross and the FEMA. The Ashland response was profoundly affected by Ashland Oil Company's decision to assume financial responsibility for the cleanup. Considering the pivotal position of Ashland officials for making decisions which affected all aspects of the response, it may have been useful to include an Ashland representative in RRT teleconferences. Their representation could have provided the RRT directly with any factual details regarding their activities and their ability to comply with RRT recommendations to the OSC. This representative would not, however, have had any authority to speak for the RRT. Coordination of Water Monitoring Efforts The coordination and communication of monitoring efforts in the initial days suffered because no lead agency was assigned. Following the initial emergency response lead by the USCG, ORSANCO accepted the lead; ORSANCO sent its field operations coordinator to the Wheeling interim command center and its Cincinnati office served as the data clearinghouse. Because there was no legal basis for ORSANCO's lead role, it was possible only by an informal agreement of the agencies involved. In the future, the early establishment of a single water monitoring data coordinating agency can serve to improve the focus on communications so that efficiency and data relevance are optimized. Central command, financial and technical responsibi- lity, and a source of up-to-date information can be established so that field monitoring may be addressed at the onset of an emergency. Furthermore, the lead agency can take steps to determine the availability of resources and establish standardization of analysis among the various agencies. 11 ------- Standardization of Sampling and Analytical Data While the degree of cooperation among the agencies was high, they operated independently so that there was a lack of standardization of analysis. Naturally, it would be optimal to use identical measures and procedures among agencies for comparison of analyses. The fluorometers, for example, provide an illustration of potential problems. Tow boat and stationary fluorometers employed discrete sample analysis and were calibrated with a standard solution of naphthalene. (There is some question about standard solution degradation during the period of analysis.) The U.S. Corps of Engineers' fluorometer employed a flow-through sample analysis and was not calibrated with a standard solution of naphthalene. Because of these inconsistencies, there was some difficulty in combining measurements to gain a complete history of the plume. As another example, some laboratories reported a nonspecific "fuel oil" concentration which could be quantified in at least two ways. One method measured a single constituent and estimates total concentration based upon fractional composition; another sums the concentrations of many similar components in a given range of the GC spectrum. The measurement techniques to be used for specific situa- tions should be standardized. Analytical methods for determining concentrations of contaminants should also be standardized. The Organics Detection System (ODS) gas chromatographs were invaluable in providing on-site analysis with minimal sample turnaround time. GC/MS was used for verification of raw and finished water quality and was required by some regulatory agencies before approval was given to water plants to return to operation. Consideration should be given to standardization of portable GCs with subsequent GC/MS verification for incidents of this type in the future. COMMUNICATIONS Communications during the initial stages of the incident were very difficult, and the fact that the incident required a shutdown of power and communications on site for a period contributed to this problem. Regardless, the sheer number of agencies involved in an incident of this magnitude would strain any communications system. eoiMnnn ications Resources During the Ashland Oil Spill, there were some problems encountered in attempting to contact RRT members during off-duty hours. A list of 24-hour phone numbers for contacting RRT members should be regularly updated and made available to all RRT 12 ------- members at all times. In addition, the 24-hour phone numbers for various agencies and the numbers to newly installed or rented portable phones need to be gathered and aggressively disseminated to all responding groups. Responders need large quantities of communications hard- ware. Immediate installation of a large number of phone lines to the command post is essential for adequate communications during a response of this size. However, cellular phones are a necessity for rapid, efficient response. Temporary rental of these phones is possible through some car rental companies. Mobile phones should be provided to field work crews so that they can make regular updates to the command post. Cellular phone use can become a problem if system capacity is exceeded due to the tremendous number of phones employed in a major disaster. Conventional phone service should be installed on scene as soon as possible. The numerous difficulties with both the EPA and USCG-NRC teleconference systems were apparent to all who participated in the RRT teleconferences. It is recommended that for future activations, the RRT should, where time permits, utilize a private telecommunications system (such as AT&T) until the federal teleconference network can be improved. EPA's and NOAA's electronic mail systems can be efficient mechanisms for communication between RRT members. It is suggest- ed that each RRT member be assigned an E-mail box under each system and that for EPA-lead responses an RRT E-mail distribution system be established at the Regional Response Center in Philadelphia. Through such a distribution system, POLREPs and Special Bulletins can be promptly delivered to all RRT members. This would improve the quality of RRT teleconferences, as all members would be in possession of the same information. The Communications Work Group for the Standing RRT should attempt to standardize the RRT E-Mail distribution system. River monitoring data were distributed using ORSANCO's electronic bulletin board and direct telephone. This system, which had been established prior to the spill with water utilities along the river, was widely used and worked well. Some parties lacked the appropriate hardware (computer and modem) to use the bulletin board and in some cases did not have knowledge of the service. Procedures could be developed to utilize ORSANCO's electronic bulletin board more effectively, or alternately for RRT use, the data could be distributed via EPA or NOAA E-Mail. Specific times of day should be established for updating the bulletin board and two-way electronic communications should be implemented. Water suppliers could then provide daily updates on their operations via computer bulletins. A water user telephone hotline could be used in a similar manner for plants without computer equipment. 13 ------- Providing Information to the Public At one point in the response, prior to receiving official notice from the state, a local government agency suggested that the need for water conservation was lessening. Criteria and authority for lifting Water Conservation Orders should be made clear to local governments and water purveyors by the party establishing the order so there is basic agreement beforehand on when conservation can be discontinued. Partial lifting of an area-wide need for conservation could confuse the public and result in prolonged water shortage in some areas. For the water suppliers, public relations activities were a high priority, essential for maintaining customer confidence in their drinking water quality. While this was generally achieved at Ashland, it is worth reiterating that information should be given in future emergencies to the media consistently and on a regular basis, preferably through one spokesperson at the facility. It is also suggested that copies of press releases be sent to other water users along the river. Very special attention must be paid to ensure that the media continually notify the public if the problem is one of quantity and not of contamination. Water suppliers should be aware that the use of interconnects, changes in water flow patterns, and varying pressures may lead to taste and odor problems that the public could misinterpret as contamination. Communications Between Command Post and Water Suppliers To the extent practicable, provisions were made so that information on the extent of contamination of raw water sources was provided to all parties involved with water supply decisions. Furthermore, modeling was used to assist in determining when contamination would reach a water intake and how long surface intakes would be shut down. Such information was, of course, vital for deciding whether or not sufficient water storage existed and which areas required emergency supplies first. More guidance is needed from state drinking water agencies in developing treatment schemes. The best information was generated with jar tests on contaminated river water samples, but it was difficult for downstream water users to obtain these samples. Consideration should be given to developing a means for delivering such samples to downstream suppliers early in the response. More timely health effects data on the contamination is needed along with assistance in interpreting its significance. Water utilities were often questioned by the public and the media about the safety of their drinking water and the toxicity of fuel oil components. In the future, ATSDR, along with state health departments, could be better utilized for this purpose on site. 14 ------- Coordination and Communication of River Monitoring Data A greater understanding of river flow is necessary for improved time-of-arrival and contaminant concentration estimates. Currently, for the Ohio River Basin, these estimates can be based only on river velocity measurements at discrete locations as provided by the River Forecast Center of the National Weather Service (NWS). In the case where the contaminant is mixed throughout the water column, as was the case with the Ashland spill, river velocity equals plume velocity. As NWS river velocity data are provided daily throughout the year for points along the Ohio River and on major tributaries, it appears an ideal source of timely and reliable information. Nevertheless, because the data are provided at wide distance intervals (e.g., 62 miles between Moundsville and Willow Island) and river velocities may change significantly in a day, considerable uncertainty exists in time- of-arrival estimates. The velocity gradient within pools is an especially important factor at low flow, when the river may take several days to travel between dams. These shortcomings were apparent in the discrepancies noted between predicted and actual times of arrival. Downstream users were kept informed of the most recent measured concentrations of contaminant and its expected treat- ability. As discussed earlier, this information was initially slow in coming because of difficulty in setting up GC analytical capability. In the future, improvements should be sought in laboratory setup procedures and by use of established water quality models. A final difficulty in the area of coordination and communication was the selection of data that would be most useful to downstream water utilities. The data first posted on the electronic bulletin board were concentrations of organic compounds, which were later supplemented by No. 2 fuel oil results. Downstream utilities remarked that this data was not meaningful to them in assessing treatability. The delivery of 5- gallon samples that could be used for treatability tests was much more useful to the water supply plants. In the end, monitoring efforts have been successful in that there was no (known) public health impact and no damage to water treatment systems due to untimely notice. RESOURCES A lack of immediately available resources such as contain- ment and monitoring equipment hindered the response to the Ashland oil spill. Federal response equipment is housed in widely separated locations and so was not easily accessible for use. No current inventories of equipment and resources 15 ------- controlled by governmental or industrial entities existed. Delays were caused by the need to locate and transport needed equipment and by the necessity of diverting valuable personnel from actual response and cleanup efforts to do so. As an example, considerable difficulty was encountered in obtaining fluorometers for spill monitoring. This tool was identified within the first days of the emergency as a means of tracking the spill, and by chance one was found with the proper light wavelength filters for oil-in-water analysis. Unfortunately, when it was decided to use the instruments in multiple locations (mobile and stationary), no information was available on other instrument suppliers. Eventually, several more fluorometers were located with the aid of the manufacturer's sales records, but by then several days' worth of data had not been obtained. A warehouse of available containment and monitoring equipment in the Pittsburgh area would have been ideal. However, the warehousing of equipment for emergency responses is not mandated by any local, state, or federal agency; nor is it cost- effective. Inventories of locally available equipment would have assisted emergency responders in quickly locating essential equipment. Such inventories could be developed through a cooperative arrangement between local industries and government. For the most part, in the initial days the regulatory agencies involved in the response were not able to provide the turnaround time for water analyses needed to make the decisions facing them. This problem may have been due to the heavy workloads or to the physical separation between the river and the laboratory. In any case, it is not likely that the analytical data could have been obtained any faster. Later, with the availability of the mobile laboratory, this problem was appreciably reduced in magnitude. Delays, of course, were especially bothersome for water utilities in need of qualitative and quantitative data to assess treatability. The ability to provide this information either through private laboratory contracts or through mobile analytical capabilities is very important to the efficiency of the response. Regulatory agencies should consider maintaining in state or local contingency plans a list of local laboratories certified to perform necessary testing in an emergency. Sampling and analysis protocols should be pre-approved by the appropriate regulatory agencies. Time estimates should be obtained from local laboratories with respect to sample processing as a part of these plans. Whenever possible, a contract with the local laboratory should be drawn up that gives emergency analyses priority over their routine work. Operating hours and number of samples that can be handled by the local laboratory should be maintained in this information package. Finally, the development of a mobile laboratory capability by the responding agencies is significant in ensuring an efficient response. 16 ------- This particular incident could have been far more devastating had public water supplies been contaminated or water shortages more severe. It is essential that emergency planning agencies and water suppliers plan for the availability of contingency water and equipment. Emergency planning agencies should maintain a current list of available bulk water haulers, facilities with tankers that can be used in refill operations (i.e., fire companies and dairies), and sources of plumbing expertise for distribution station hookups. Each water supplier should maintain a list of service and equipment companies that can provide replacement pumps, chlorination equipment, or chemical feed equipment in the event of an emergency. 17 ------- SECTION IV CONCLUSION ------- SECTION IV CONCLUSION In particular, the RRT notes that the willing cooperation between all regional elements was the outstandng factor that resulted in the successful protection of the public health and the environment during the Ashland Oil Spill. All responding agencies, groups, and individuals are to be commended for their performance throughout the emergency. Despite the magnitude of the spill, the rapid entry of oil into the river, and adverse weather conditions, all public water suppliers were protected. The RRT concludes that the overall response to the Ashland Oil Spill was effective. Despite the magnitude and instantaneous nature of the oil release and the various factors, including weather conditions, that hampered containment and cleanup efforts, the response achieved the following: a) The initial notification of the spill was timely; b) River monitoring efforts were successful in that drinking water intakes were secured prior to any contamination of water supply systems; c) Ashland Oil Company worked cooperatively under the direction of the U.S. EPA and agreed to take full responsibility for funding cleanup activities; and d) The spill site was safely secured with no injuries to workers, emergency response personnel, or residents. Although the response to the Ashland Oil Spill was effective, the RRT's final conclusion is that implementation of the recommendations presented herein will improve the efficiency and effectiveness of future responses. 18 ------- APPENDIX A WATER USER EXPERIENCE ------- APPENDIX A WATER USER EXPERIENCE Water supply plants downstream from the Ashland terminal were promptly notified of the January 2, 1988 spill; however, the presence of oil throughout the water column and the corresponding impact on water supplies was not recognized until the following morning. This Appendix summarizes the experiences of the major water supply facilities that were affected by the Ashland Oil Spill. West Penn Water Company (Becks Run^. PA fMononaahela River mile 4.5^ West Penn Water (WPW) was notified by the Allegheny County Health Department (ACHD) about the spill at approximately 7:00 p.m., January 2. Little information was available on the spill location, size, volume (estimates ranged from 30,000 to 3 million gallons), or travel time to intake. The Hays Mine plant intake was closed early January 3, and testing began immediately using the ORSANCO-ODS gas chromatograph. Additional samples were sent to a local lab for GC/MS verification. Samples were also taken by EPA, PADER, and ACHD. Attempts were made to determine the chemical composition of the fuel oil. The plant considered blocking off the two upper water intakes and operating with the lower intakes; however, oil was observed at depths greater than 15 feet. On Monday, January 4, WPW pilot tested a carbon feed process that proved successful in removing the oil. This process was used by four of the five Pennsylvania water companies that used the contaminated portions of the river as a raw water source. The fifth water company, Sewickley Borough, switched over to wells. On January 6, treatment and analysis schedules were established and the Hays Mine plant started operating at 25 percent capacity. Excellent public relations efforts helped to maintain public confidence throughout the water emergency. City of Pittsburgh Water Authority. PA (Allegheny River mile 7.41 The City of Pittsburgh draws its water from the Allegheny River, which was uncontaminated by the spill. The excess capacity of the Pittsburgh system provided 15 to 20 million gallons per day (mgd) to West Penn and West View Water Companies. ------- West View Water Authority. PA (Ohio River mile 4.5) West View Water Authority was notified of the spill the evening of January 2, and immediately implemented their emergency contingency plan. This plan included increasing carbon feed capacity, topping off reservoirs and tanks in the system, activating wellfields, and establishing water line connections with the City of Pittsburgh. Use of the developed wells was delayed for a short time pending laboratory evaluation of the quality of the raw water. Regulatory agencies approved operation on January 10, at which time the plant started to blend 25 percent river water with well water. The primary factors that enabled West View to provide water to its customers were the use of well water, the connections with the City of Pittsburgh's system, and the Pennsylvania Emergency Management Council's conservation order. Robinson Township Authority. PA (Ohio River mile 8.6) Robinson Township was the only water company that had customers without water during the incident. Approximately 17,000 customers were without water for up to 48 hours. Two thousand (2,000) customers were without water for up to 5 days, and 200 customers were without water for an entire week. The primary activity of the Authority during the incident was the development of interconnections with other water supply plants. The plant was able to open its river intakes 8 days after the spill, on January 10, following conversion from a potassium permanganate feed system to a carbon feed system. Midland Water Authority. PA (Ohio River mile 36.2) Midland Water Authority was advised on January 3 at approximately 7:00 p.m. that the oil would reach the plant in 24 hours, and were advised to build reserves and cut consumption. The river intake was shut down at 6:00 p.m. on January 4, with 4 million gallons of water in reserve. No interconnections with other water supply plants were possible. A local industry, J&L Products, which normally uses 2 MGD, was ordered by PADER to cut consumption to a minimum. On January 7, divers completed the installation of oil filter blanket material around the intakes, and plant startup procedures began using 600-pound/million gallons powdered activated charcoal for water treatment. Results of tests of the finished water were acceptable and the plant went back on line with the modified treatment process at 2:00 p.m. on January 7, 1988. ------- East Liverpool Water Treatment Plant. OH (Ohio River mile 41) East Liverpool water treatment plant shut its intakes at 0300 hours on January 5, 1988 even though the plant had not yet been affected by the spill. Officials of the plant determined that they could not afford the risk of contaminating the facility. In the interim, reserves and water conservation were relied upon until river water treatment could commence. A creek outflow aided in minimizing the spill's effects by diverting oil from collecting near the facility's intake. By 1200 hours on January 7, East Liverpool was back on line treating water with activated carbon. Ashland's mobile laboratory was utilized to test water samples on a 24-hour basis to ensure water quality. Toronto Water Treatment Plant. OH (Ohio River mile 60) At 2300 hours on January 5, 1988, Toronto, Ohio closed its intakes. A large reservoir lessened the severity of the situation. The reservoir, coupled with water conservation efforts, provided sufficient water supplies until the plant reopened its intakes and began treating river water at 2230 hours on January 9, 1988. Steubenville Water Treatment Plant. OH (Ohio River mile 65) At the time of the spill, Steubenville's storage capacity was at its fullest enabling an adequate supply of water to be provided. Facility officials monitored the quality of water at 15 to 20-minute intervals. When odor became apparent on January 6, 1988, the water intake was closed for a period of 10 to 11 hours. After the contamination had passed, the intake was reopened at 2225 hours. Wheeling Water Treatment Plant. WV (Ohio River mile 86.8) Useful information on treatment methods was developed by water utilities upstream. The plant used a number of tests to monitor for oil in water, such as the Hanby Test, fluorometric analysis, threshhold odor, and gas chromatography. The GC was used to confirm the presence or absence of low-level organics in raw and finished water. Treatment schemes were evaluated using jars of river water spiked with oil. Reliable concentration and health effects data from upstream monitoring were not readily available during the first week following the spill. Alternate water sources were obtained from the communities of Martins Ferry and Bridgeport and water barges brought in by Ashland Oil. These barges were filled with water from Wheeling Creek, a tributary of the Ohio River downstream of the Wheeling water treatment plant. The plant went off Ohio River water the evening of January 8 and reopened by blending with barge water on January 10. Prior to reopening, an absorbent filter blanket was placed over the intakes by divers. ------- American Electric Power. WV (Ohio River mile 76.5. 111.1. lii.su Power plants utilizing the river for cooling water are vulnerable to high levels of oil. Since the period immediately after the spill saw near record demands for electricity due to heating needs, the threat of disruption of any of the power plants was a grave concern. The fact that there was no disrup- tion is noteworthy; however, plans for water treatment, if it became necessary, were in place. The Cardinal Plant, at mile 76.5, was the only plant that observed oil at the 20 ppm level. A low dose of detergent was added to the water there. Sistersville Water Works. WV (Ohio River mile 1371 There was much concern regarding potential difficulties in providing adequate water supplies at Sistersville because there were no alternate supplies. During the crisis, Ashland provided barged water to Sistersville while the plant's intakes were closed from January 13 to January 21, 1988. USCG assistance was not required. Residents also assisted by adhering to water conservation requests. E.I. DuPont. Parkersbura. WV (Ohio River mile 1901 The oil spill had no immediate impact on plant operations. Huntington Water Treatment Plant. WV (Ohio River mile 3061 Huntington was hampered by a water hysteria in fear of a water shortage prior to the arrival of the spill to the area. Diesel pumps were installed in a nearby, unaffected stream as an alternative water supply in anticipation of potential river contamination. Intakes were closed on January 22 and water from the Guyandotte and Ohio Rivers was blended for treatment until 100% Ohio River water treatment resumed on January 25, 1988. Ashland Water Treatment Plant. KY (Ohio River mile 319.75) Ashland, Kentucky shut its intakes at 0200 hours on January 22 and reopened them and began treating river water at 1200 hours the following day. In the interim, reserves were utilized. Mavsville Utility Commission. KY (Ohio River mile 408.5) Barged water was provided to Maysville by Ashland Oil Company during the period from 0715 hours on January 23 until 1630 hours January 24, 1988 while the intakes were closed. ------- Cincinnati Water Works. OH (Ohio River mile 462.8) The distance of Cincinnati from the spill site allowed more time for implementation of contingency plans. Information about the spill's concentration as it moved downriver was limited and was not helpful in predicting treatment dosages of activated carbon. Plant personnel traveled to Wheeling and Huntington to collect samples of the oil for jar test analyses. Fluorometers were used for analysis of samples and provided better estimates of the oil concentration. GC/FID was the most sensitive tool used for organics detection; however, the fluorometer was easier to operate and provided a faster turnaround time on test results. Threshhold odor was not an effective test at the low oil contaminant levels observed at the Cincinnati Water Works. The plant remained off the river water for 86 hours and resumed operation with carbon treatment until no trace of oil contamination was observed in the river. Kenton County Water District No. 1. KY (Ohio River mile 463) On January 21, the Kentucky Department of Natural Resources held a meeting with water utility personnel and state emergency teams to discuss response plans when the spill reached the Kentucky border. Information on the spill's position and profile was presented, monitoring procedures were established, and a communications network set up. The Kenton County plant was off the Ohio River for about 72 hours and no serious problems were encountered. Since 2 sources were available, either the Ohio and the Licking Rivers, the plant minimized the amount of water treated with activated carbon. Louisville Water Company. KY (Ohio River mile 600.6) The Louisville Water Company (LWC) found the upstream water user information very helpful and relied on ORSANCO to provide updates on spill movement and time of arrival estimates. As the spill diluted and spread downriver, it became apparent that LWC would be faced with low levels of oil contamination over a long period of time. Therefore, the company focused on the develop- ment of treatment strategies and on keeping the public informed about the steps taken. Threshhold odor, fluorometry, and gas chromatography were among the tests used to monitor the oil contamination. Odor tests were conducted using water spiked with No. 2 fuel oil. Five (5) ppb appeared to be the minimum detection level for the odor test. Various treatment schemes were evaluated with jar tests using oil concentrations of 100 ppm to 10 ppb. ------- The first oil contamination was observed in the raw water at noon on January 27 and oil reached a maximum level of 67 ppb on January 29. Activated carbon was added (between 200 and 400- pound/million gallons) during this period and was gradually decreased as the contamination dropped below 1 ppb 4 days later. Evansville Water Works. IN (Ohio River mile 791.5) Treatment information and concentration data obtained from upstream water users was sufficient for contingency planning at Evansville. River conditions changed drastically with heavy rainfall the week of February 1. This reduced the time-of- arrival estimates by 2 weeks. The spill's leading edge was no longer distinct by the time the spill arrived and fluorometer peak readings were inconsistent. Fluorometry and gas chromatography were used to monitor raw and finished water. The passage of the oil took approximately 20 hours, with a 10 ppb oil concentration at the peak. Carbon treatment was applied for 5 days. Media coverage and public communication required a substantial amount of time throughout the month of January. Paducah Water Works. KY (Ohio River mile 935.5) Since the spill had diluted beyond detection by the time it reached Paducah, no additional treatment measures were taken at the plant. Media coverage, however, was intense and public relations activities consumed a great deal of time. Monitoring at water plant intakes had been planned by Kentucky as long as the oil contamination was detectable. Paducah was not informed when oil monitoring was discontinued by the state. ------- APPENDIX B TRANSCRIPT OF THE NRC REPORT OF THE ASHLAND OIL SPILL FLOREFFE, ALLEGHENY COUNTY, PA ------- TRANSCRIPT OF NATIONAL RESPONSE CENTER TAPE REPORT OF THE ASHLAND MAJOR OIL SPILL FLOREFFE, ALLEGHENY COUNTY, PENNSYLVANIA It should be noted prior to reading this text that during the course of this report, date and time were not given. It is known that the date of the initial spill report was January 2, 1988; however, the exact time of these events is not known. Call #1 • NRC Incoming NRC: National Response Center. Caller: Yes, sir. This is John Gajdosik at Floreffe, Pennsylvania. I have an emergency. 1 have a diesel fuel tank that has collapsed. NRC: All right, sir, you know that as far as an emergency situation, all I can do is take a report in a timely manner and call the federal on-scene coordinator. Caller: Okay, well, I'm going to be calling the Coast Guard, in a second. You are the National Response Center, so 1 wanted to notify you. NRC: Well, let me get a report form started. Okay? Caller: Okay. NRC: I've got a series of questions to ask you. Pause. Okay, your name again, sir, is what? Caller: John, that's G-a-j-d-o-s-i-k. Look, I'm going to have to go shut off some . . . I'm going to have to go shut off some electricity. NRC: Let me, let me have your phone number right quick. Caller: Okay. That's area code 412, 462-5111. NRC: And you're in what city? Caller: Floreffe, Pennsylvania. That's south of Pittsburgh. NRC: Uhat county? Caller: A11egheny. NRC: And the location where you're at? Caller: Uh. You mean the, where do you mean where am I at? NRC: Where the tank's at. Caller: Uh. It's inside the terminal at Ashland Oil, Floreffe. ------- NRC T rartscr i pt Page 2 NRC: Ashland Oil? Caller: Ashland Oil Company. NRC: And the street is . . . Caller: State Street, on Route 837. NRC: Okay, sir. Caller: Okay? NRC: All right. Caller: All right, bye. NRC: Bye. Call #2 - NRC Outgoing Answer: Hello. NRC: Hello. This is Petty Officer Mackey at the National Response Center. Answer: Yeah. NRC: I got a real sketchy report from--are you familiar with Ashland Oil on State Street, Route 837 in, he called it Floral, PA? He said it was just south of Pittsburgh? Answer: Floreffe. NRC: Floreffe. Answer: I know where it's at. NRC: Okay. He was out of breath when he called me, so all I got was that and his phone number. He said he had a diesel tank that collapsed. And he had to call the Coast Guard. Answer: He had a diesel tank? NRC: A diesel tank that collapsed. And he told me he couldn't document any more information, that he had to go close some valves, so he was, um Answer: All right, what's his phone number there. ------- NRC Transcript Page 3 NRC: Area code 412, 462-5111. And I'm going to give this NRC Report #3 2. Answer: Number 32? NRC: Uh-huh. His name was Gados i k, G-a-d-o-s-i-k. Answer: That's all right, I can get it. NRC: Okay, and I hope you'll call me back and I'll finish my report. Answer: I'll give you a call back. NRC: Okay. Answer: That'll make it easy. NRC: That's in Pittsburgh's area? I got . . . Answer: Yeah, it's mile, um--oh--25 on the Monongahela River. NRC: Okay. Answer: Give or take. NRC: Well, it sounds like you got a pretty . . . Answer: Big one NRC : Big one goi n' on. Answer: Did he say any of it was getting in the water, or what? NRC: Ah, he never mentioned that. He said that he was wanting to call the Coast Guard. He said he wanted to notify the National Response Center and then he wanted the Coast Guard to be notified right away. So . . . Answer: Evidently it was. Tank collapse. He didn't say if it was a.tank barge or if it was . . . NRC: Uh-uh. He said a tank, so I think it's a storage tank. They have any tank--a tank farm there? Answer: Yeah. NRC: That's what I got the feeling that it is. Answer: They got a small tank farm there. Okay. Appreciate it. NRC: All right. Bye. Answer: Bye. ------- NRC T ranscri pt Page 4 Call #3 - NRC Incoming NRC: National Response Center. Caller: Yes. I'd like to report a spill, please. NRC: All right, just a moment, sir. Pause NRC: Before you get started--] was talking with one guy--you're not calling from Ashland Oil, are you? Yes I am. All right. Is this in Pennsylvania? Right. All right. So I can just reference the same report that--'cause I got started . . . Who were you talking to there? Nagadosik?--or? G a d o s i k ? Yeah. Okay. That's the right way to spell his name-•G•a-d-o-s-i-k? G-a-d-j-o-s-i-k'-yeah. G-a-d-j . G-a-j-d-o-s-i-k. CaI ler: NRC: Caller: NRC : Caller: NRC: Ca I ler: NRC: Caller: NRC: Caller: NRC: Caller: Long pause NRC: Caller: NRC: Okay, that report number's gonna representing Ashland? Ash I and Oil, right. Are you representing-•you' Well, I'm gonna put your name in place of h i s • - what was your name, sir? Caller: Morgan. ------- NRC Transcript Page 5 NRC: Morgan? Caller: Yes. M-o-r-g-a-n. NRC: And your first name? Ca11er: George. NRC: And you're at 412-462-5111? Caller: They're out of service right now, but yes, that is our number. NRC: Uhat's another number? CaI Ier: Pardon? NRC: Is there another number? Caller: Right now we don't have one. I'm over at the neighboring factory here. Ah, let's see, 384-3350. NRC: 384? Caller: 3350. f NRC: And it is--the proper name of the company is Ashland Oil? Caller: Ashland, yes. Ashland Oil, Inc. NRC: The address is State Street and Route 837? Caller: Uh--204 Glass House Road. NRC:, And that's Floreffe, F-1•o-r•e-f-f-e, Pennsylvania? Caller: Right. 15025. NRC: 15025. Caller: Right. NRC: And that's mile 25 on the Nongona--Nongahelia River. Caller: Honongahela River, right. NRC: Allegheny County. Caller: Right. ------- NRC Transcript Page 6 NRC: Spill occurred at what time? Caller: I was notified at 5:30, so I imagine sometime between 5 and 5:30. NRC: Because he called me at 17:22. Caller: That's about it then, because he called me probably about the same time. NRC: All right. Spilled diesel fuel. Do we have an amount? Caller: Ah--an undetermined amount right now, but the tank ruptured, so it hard to say. NRC: And it's a storage tank? Caller: Right. MRC: And it collapsed. Caller: Right. NRC: Is it going into the water? Caller: 1 haven't down been on the river, but I've heard that it is, so I' assuming that it is. NRC: And, what's being done? Do you have any idea? Caller: Right now we have all the local fire departments; they got everything cordoned off. They're bringing in some sand trucks and some end loaders to put out on Route 837. And we're trying to survey the situation right now over in the terminal to see what we have got there exactly. NRC: Okay. What was the capacity of this tank? Caller: Ah, 90,000 barrels. NRC: So it's potential that 90,000 barrels have spilled out. Caller: I--the max I would say, if indeed it's true, would be about 80,000 If we've lost it all, I haven't been able to determine that yet. NRC: Have you notified any other agencies? Caller: Ah, just our own emergency group and we'll be calling the Coast Guard ... NRC: I've notified the Coast Guard in Pittsburgh, so . . . ------- NRC Transcript Page 7 Caller: Okay. NRC: ... a I ready. Caller: Okay, that'll take care of that, then. NRC: And, I was looking at any state agencies yet. Caller: Ah, no I haven't yet, that'll be the next one on the line then. NRC: That's the Pennsylvania DNR? Caller: Yeah. NRC: All right, so it looks like a pretty bad situation. Caller: 1 would say right now it's a very bad situation. NRC: All right. Caller: Okay, could I have your name please? NRC: It's Petty Officer Mac key *•M-a•c•k-e-y. Caller: H-a-c-k-e-y. Thank you. NRC: Yes, sir. Caller: Bye. Call #4 • NRC Incoming NRC: National Response Center. Caller: Yes, sir. My name is James Smith. I'm the safety office, Hazardous Materials Response Team from the PIeasantvi I I e Volunteer Fire Company. Ue are currently at the scene of a diesel fuel storage tank coll apse. NRC: Yes, s i r - • i n Floreffe, Pennsylvania. Caller: Yes, sir. You've been made aware of that? NRC: Yes, sir. Do you have any details on how much has been spilled? ------- NRC Transcript Page 8 Caller: Yes, sir. Ah, the best details we can get at the moment are approximately 100,000 gallons. Presently, the containment dike surrounding the tank has contained most of it. Unknown as to how much. We do have a spill across the highway and some towards the river, however we are unable to ascertain exactly (faded). We have contained it as best as possible. In the process of containing it to the best of our ability. Local emergency management has been notified, but we definitely do have approximately 100,000 gallons. NRC: All right. But not that much in the river. Catler:No,sir. NRC: All right. Well, I've just talked with the owner and he said that it's a 90,000-barrel capacity tank. Ca I I e r : Okay. NRC: And he said at the most he suspect 80,000 barrels spilled. That's a whole lot more than 100,000 gallons. And, ah--but the tank is surrounded with a dike? Caller: Yes, sir. NRC: And the dike?--it's overflowing the dike? Caller: Yes, sir. NRC: All right. Ca 11 er: Okay? NRC: I'll put that into the report that I got from them. Caller: Okay, thank you, sir. NRC: Bye. ------- APPENDIX C INCIDENT-SPECIFIC REGIONAL RESPONSE TEAM ------- APPENDIX C INCIDENT-SPECIFIC REGIONAL RESPONSE TEAM ASHLAND MAJOR OIL SPILL THOMAS C. VOLTAGGIO, CO-CHAIRMAN DENNIS CARNEY ROBERT E. CARON STEPHEN D. JARVELA ALAN JACKSON U.S. Environmental Protection Agency Region III, Superfund Branch 841 Chestnut Building Philadelphia, PA 19107 (215) 597-9893 FRED STROUD U.S. Environmental Protection Agency Region IV 345 Courtland Street, N.E. Atlanta, GA 30365 (404) 347-3931 MICHAEL STRIMBU U.S. Environmental Protection Agency Region V 230 S. Dearborne Street Chicago, IL 60604 (312) 353-2102 CMDR JAMES CLOW U.S. Coast Guard, Fifth District (meps) 431 Crawford Street Portsmouth, VA 23705 (804) 398-6638 CMDR ROBERT LUCHUN U.S. Coast Guard, Second District (meps) 1430 Olive Street St. Louis, MO 63103-2398 CMDR EUGENE A. MIKLAUCIC U.S. Coast Guard, Marine Safety Office Second District Suite 700, Kossman Building Forbes Avenue and Stanwix Street Pittsburgh, PA 15222 (412) 644-5808 ------- ANITA MILLER U.S. Department of Interior U.S. Courthouse, Room 502 Second and Chestnut Streets Philadelphia, PA 19106 (215) 597-5378 DR. JEAN SNIDER U.S. Department of Commerce National Oceanic and Atmospheric Administration Hazardous Material Response Branch 11400 Rockville Pike Rockville, MD 20852 (301) 443-8933 JOSEPH McCAREY Federal Emergency Management Agency 105 South 7th Street Philadelphia, PA 19106 (215) 931-5520 COL MATT MILLER U.S. Army Corps of Engineers, Pittsburgh District Hydrology and Hydraulics Branch 1928 Federal Building, 1000 Liberty Avenue Pittsburgh, PA 15222 (412) 644-6831 JOHN BARRY U.S. Department of Labor Occupational Safety and Health Administration Gateway Building, Suite 2100 3535 Market Street Philadelphia, PA 19104 (215) 596-1201 PETER TENNANT ORSANCO 49 East 4th Street Suite 815 Cincinnati, OH 45202 (513) 421-1151 FRED OSMAN Pennsylvania Department of Environmental Resources P.O. Box 2063, Fulton Building Harrisburg, PA 17120 (717) 787-2814, 5028 JOSEPH LaFLEUR Pennsylvania Emergency Management Agency P.O. Box 3321 Harrisburg, PA 17105 (717) 783-8016 ------- RON SANDY West Virginia Department of Natural Resources 1201 Greenbrier Street Charleston, WV 25311 (304) 348-3614 RUSS RADER Assistant: Director, Environmental Engineering Division West Virginia Department of Health 1800 Washington Street E., Room 550 Charleston, WV 25305 (304) 348-2981 WILLIAM BURGER Kentucky Department for Environmental Protection Division of Water Environmental Response Team 18 Reilly Road Frankfort, KY 40601 (502) 564-3410 ZACK CLAYTON Ohio Environmental Protection Agency 1800 Watermark Drive P.O. Box 1049 Columbus, OH 43266-0149 (614) 481-4300, ------- APPENDIX D RIVER MONITORING WORK GROUP ------- APPENDIX D RIVER MONITORING WORK GROUP FEBRUARY 11, 1988 SEWICKLEY , PENNSYLVANIA Participants in the Hork Group to conduct the critique of the initial response at the Ashland Major Oil Spill was chaired by ORSANCO included representatives of the agencies which were involved in the monitoring efforts and included: NAME Scott Fennell ORGANIZATION ORSANCO ADDRESS 49 E. 4th Street Cincinnati, OH 45202 TELEPHONE NO. (513) 421-1151 Russ Rader John Youger UV Health Dept Ohio EPA Charleston, UV 25311 1800 Uatermark Drive Columbus, OH 43215 (304) 348-2981 (614) 481-7130 Chuck Taylor Ohio EPA 1800 Uatermark Drive Columbus, OH 43215 (614) 481-7025 Steve Steranchak PADER 121 S. Highland Ave. Pittsburgh, PA 15206 (412) 645-7100 Patricia Miller PADER 121 S. Highland Ave. Pittsburgh, PA 15206 (412) 645-7100 Russ Stotzman Tom Proch Mike Havelka Bob Moran Peggy Uallace PADER PADER EPA/Ueston TAT Indiana DEM CHMR 121 S. Highland Ave. Pittsburgh, PA 15206 121 S. Highland Ave. Pittsburgh, PA 15206 Suite 436, Hawley Bldg Wheeling, UV 26003 5500 Bradbury Indianapolis, IN 46241 320 Uilliam Pitt Uay Pittsburgh, PA 15238 (412) 645-7100 (412) 645-7100 (304) 233-1610 (317) 342-5158 (412) 826-5320 Sarah Shockley CHMR UPARC (412) 826-5320 Lythia Metzmeier KY Div. of Water 18 Reilly Road (502) 564-3410 Frankfort, KY Ron Sandy WVDNR 1201 Greenbrier (304) 348-3614 Charleston, WV ------- NAME Sam Perris Nark Anthony Gerald L. Greiner James Amnion Patr i ck Neichter George Kincaid Michael Koryak Bill Cremeans Robert W. Schmitt Jim Irwin Michael Dalton Gene M i kIauc i c Jerry Schulte Peter Tennant ORGANIZATION WVDNR Corps of Engineers PA Fish Commission PA Fish Commission Corps of Engineers Corps of Engineers. Corps of Engineers Corps of Engineers Corps of Engineers Ohio EPA Ohio EPA U.S. Coast Guard ORSANCO ORSANCO ADDRESS 1304 Goose River Rd. Fairmont, WV Cincinnati, OH 1520 Menk Road New Kensington, PA P.O. Box 31 U. Mifflin, PA P.O. Box 59 Louisvi11e, KY 40201 Huntington, UV Pittsburgh, PA Huntington, WV Pittsburgh, PA Twinsburg, OH 1800 Watermark Drive Columbus, OH 43215 Pittsburgh, PA 49 E . 4th St. Cincinnati, OH 45202 49 E. 4th St. Cincinnati, OH 45202 TELEPHONE NO. (304) 366-5880 (513) 684-3070 (412) 339-1564 (412) 672-0740 (502) 582-6739 (304) 529-5694 (412) 644-6831 (304) 529-5338 (412) 644-6951 (216) 425-9171 (614) 481-4300 (412) 644-5808 (513) 421-1151 (513) 421-1151 ------- APPENDIX E GLOSSARY OF ABBREVIATIONS ------- APPENDIX E GLOSSARY OF ABBREVIATIONS ATSDR Agency for Toxic Substances and Disease Registry COE United States Army Corps of Engineers DOI United States Department of Interior EPA United States Environmental Protection Agency FEMA Federal Emergency Management Agency IDEM Indiana Department of Environmental Management KDW Kentucky Division of Water NOAA National Oceanic and Atmospheric Administration NRC National Response Center OEPA Ohio Environmental Protection Agency ORSANCO Ohio River Valley Water Sanitation Commission OSC On-Scene Coordinator PADER Pennsylvania Department of Environmental Resources PEMA Pennsylvania Emergency Management Agency PFC Pennsylvania Fish Commission RRT EPA Regional Response Team TATM EPA Technical Assistance Team Member USCG United States Coast Guard WVDNR West Virginia Department of Natural Resources ------- APPENDIX F ASSESSMENT OF MONITORING ELEMENTS ------- APPENDIX F ASSESSMENT OF MONITORING ELEMENTS For purposes of assessment, monitoring efforts in the after- math of the spill have been divided into two categories based on the types of analyses. The categories are analyses providing qualitative data and those providing quantitative data. Qualitative Data Qualitative measures are those that provide a measure of relative concentration. This type of measurement was found to address the problem of tracking the spill in the most timely fashion. Fluorometrv The fluorometer was found to be an exceptionally useful tool in tracking the occurrence of the plume. Fluorometers relate the visible light emitted by fluorescing compounds to units of concentration. Certain components of the diesel fuel (naphthalene and benzene had been identified) made the fluorometer a viable instrument in this case. Moreover, the instrument is rugged and transportable, easy to operate and provides nearly instant results. Figure 1 shows a concentration profile as measured with the boat fluorometers. Figure 2 shows the apparent dilution of the plume as it traveled downstream. The fluorometer performed admirably in its intended use of tracking the spill. Nevertheless, there are unanswered questions regarding further interpretation of the results. Ultimately, it may be possible to convert measures of fluor- escence into diesel fuel concentration. This will require ident- ification of diesel fuel components that fluoresce at 524 nm, and the determination of their relative concentration in diesel fuel. In addition, account must be taken of the relative degradation or volatilization of the fluorescing versus non-fluorescing components with time. Turbidity, which varied greatly over the monitoring period, may affect fluorescence as well. Two areas for exploring the use of fluorometry include: 1) Its use in quantifying concentration; and 2) its utility for other contaminants. The first area will become apparent to some extent as analysis of data progresses. As indicated in the body of the report, initial analyses indicate linear correlation between fluorescence and concentration. A literature search may reveal other studies addressing the issue. Secondly, manufac- turers' literature may be helpful in identifying other compounds detected by the fluorometer. This information, along with an ------- inventory of materials commonly stored or shipped on the river, would be useful for development of emergency response monitoring programs. Hanbv Test The Hanby Test is a simple extraction/colorimetric test that may be performed with a field kit in about 15 minutes. The principal reactions involve an extraction reagent reacting with aromatics in the contaminated sample. A catalyst is then added to create color, the hue and intensity of which is compared to a standard chart for the interpretation of concentration. While the method has the desired characteristics of simplicity and immediacy, there are apparent shortcomings. Determination of concentration with the Hanby Test is a sub- jective judgement, which may not be reproducible among several field personnel. Moreover, because the reaction is subject to photochemical oxidation (i.e., in sunlight), the sample must be analyzed within minutes, and is thus unavailable for comparison with subsequent samples. According to information in the manu- facturer's brochure, the method appears to be effective up to 20 ppm of diesel fuel; analytical results from the Pittsburgh area indicate the diesel fuel concentration to be on the order of 500 ppm (Robinson Township intake, EPA sample of January 6, 1988). In light of these problems, the fluorometer was preferred over the Hanby Test for continued field use. Meanwhile, the manufacturer is pursuing improvements to the method, which may make the test viable in the future. Odor Odor as a means of determining contamination remains a standard practice at water utilities along the river, although it was not pursued in the coordinated efforts. The method is very simple and results quickly obtained. When heated, even dilute concentrations of diesel fuel (less than 1 ppm) were apparent in water-intake samples as a kerosene-like smell. For example, the laboratory technician at the Portsmouth Water Plant success- fully detected the leading edge and peak concentration arrivals by odor as later confirmed by fluorometer data. The method was not pursued in boat monitoring because a more definitive measure was desired and because continuous odor sampling is hampered by fatigue (as well as proximity to diesel engines on tow boats). Quantitative Data Quantitative measures have the advantage of providing real concentration figures which may be more useful in decisions about treatability, threats to wildlife and fisheries, etc. Unfortu- nately, measurements take more time to obtain and uncertainty of results remains a problem. ------- Total Organic Carbon Total organic carbon (TOC) is a measure of the total organic content of a contaminated water sample, including the aromatic and aliphatic compounds that make up diesel fuel. The required analysis time is relatively short (approximately 1/2 hour), but because the equipment is designed for the laboratory, it is unsuitable for field measurements. Another shortcoming is the uncertainty of the analysis. According to Standard Methods (1985), the precision of Combustion-Infrared Method is 5 to 10 percent on unfiltered samples. But a bigger problem may be the nonspecificity of the analysis. In the case of tracking a spill, there is no assurance that increased TOC is due to the occurrence of the contaminant plume versus unknown discharges. Indeed, initial review of the data appears to affirm the decision not to rely on TOC as the primary means of tracking the plume. The TOC data showed no logical trend. Oil and Grease Oil and grease (O&G) is a nonspecific measurement of the occurrence of similar compounds which dissolve in trichlorotri- fluoroethane. It includes all materials extracted from the acidified sample and not volatilized during the test, including sulfur compounds and chlorophyll. A significant fraction of diesel fuel may be left unextracted by the solvent or lost to volatilization, depending largely on method (Standard Methods, 1985, p. 497). Like TOC, it may be unsuitable for tracking a plume where interference from other contaminants complicates interpretation of the data. Gas Chromatography For identification of contaminants and quantification of concentration, gas chromatography (GC) is the preferred method. Detection limits are typically about 10 ppb or less, sufficient for purposes of addressing treatability. In an emergency, how- ever, a drawback is the relatively long analysis time ranging from one hour to one day. Also, the delicate nature of the equipment precludes field use and necessitates employment of analytical laboratories, which are costly in time and money. The most common use of gas chromatography in the affected area is for analysis of volatile organic compounds. Ten water utilities along the Monongahela and Ohio Rivers are participants in the ORSANCO Organics Detection System and collect daily samples for volatile organics analysis. The diesel fuel spilled, however, did not contain significant quantities of volatiles. GC analysis for base neutrals was necessary to identify the signifi- cant components of the material spilled. That type of analysis is more time consuming as well as less widely available than that for volatiles. At best, results were available 24 hours after sample collection. ------- For purposes of evaluating the recent monitoring efforts and developing monitoring plans for future events, it is desirable to combine or compare data obtained by different measurements. This may include, for example, correlations of gas chromatography and fluorometer data in order to improve the use of the latter. Initial analyses performed by the Huntington District U.S. Corps of Engineers indicate great potential of using fluorometry (perhaps combined with turbidity measurementss) for the determination of contaminant concentration in ppm. Similar comparisons of odor detectability, Hanby Test, etc. may reveal potential supplements to fluorometry for tracking and perhaps quantifying contamination. Finally, a study of the actual plume velocities versus NWS river velocities is also in order for the eventual improvement of time-of-arrival prediction methodology. ------- w u 2? UJ y D 3 u- FIGURE 1 PLUME PROFILE OF 15,16 JAN LEADING EDGE TO THE SPILL SITE RIVER MILE ------- 4.5 4 3 J5 3 23 2 1B 1 0.5 0 AND BACKGROUND FLUORESCENCE COMMON FLUORESCENCE UNITS ~ \ RAIN •4f T 100 300 500 RAIN 700 OHIO RIVER MILE + BACKGROUND VALUES ------- |