Feasibility Study for Highway Hazardous Materials Bulk Package Accident Performance Data Collection (2013)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

109 S e c t i o n 5 The fourth objective of HMCRP Project 07 was to identify potential institutional barriers to the development of a data- base and approaches to overcoming them. Since the database (created by an addendum to Form DOT F 5800.1) would be a government-managed activity, the associated institutional issues must be considered. These include the cost, the effect of public transparency on participation, data quality, and the regulatory hurdles of implementation. There are also institutional barriers associated with the parties that would report the accident data. These barriers stem from the per- ception that there is little benefit in reporting accident data and even some potentially negative consequences. The poten- tially negative consequences include the cost of completing reports and the risk that the information provided could be used in litigation against the provider. Additionally, it may be difficult for individuals reporting an accident to obtain damage information. This difficulty may be due to accident scene safety measures (i.e., evacuating the scene and clear- ing the roadway) and the disposition of the bulk package following an accident (i.e., being placed in an impound yard). These challenges will need to be overcome in order to achieve a successful data collection process. Industry Opinion A successful accident damage database will require collec- tion of relevant information at minimal cost and with full cooperation or participation from key industry stakeholders. Industry concerns with gathering additional accident dam- age information need to be understood and overcome. To gauge the potential level of participation from the indus- try and understand its concerns, an extensive interview and survey process was conducted with the following industry associations: • Truck Trailer Manufacturers Association (TTMA). • International Tank Container Organization (ITCO). • National Tank Truck Carriers Inc. (NTTC). • American Chemistry Council (ACC). • American Petroleum Institute (API). • Compressed Gas Association (CGA). • American Trucking Associations (ATA). • American Transportation Research Institute (ATRI). Concerns Raised by Industry Associations Several stakeholder group representatives expressed little confidence that a voluntary database could be successfully established due to the 200 to 300 stakeholders involved in the trucking industry. By comparison, the rail industry tank car accident database involves only a dozen or so companies (See Appendix C for a discussion of this and other institu- tional differences affecting the development of a database). Moreover, several of the industry associations also expressed concerns about an accident damage database. The following subsections summarize the principal concerns. Influence on Modal Competition Some industry associations expressed concern that a ship- per’s use of the database and resulting analyses could lead them to shift traffic away from truck to rail. Ironically, the hazardous material shippers interviewed believed the oppo- site, stating that if there were adequate data to reliably assess highway transport safety and risk, motor carriers would obtain a larger share of business. Shippers believe that high- way transport is less risky than it is currently perceived to be, but they must discount risk estimates because of the lack of the subject database needed to objectively assess risk. Ship- pers would prefer a level of quality in highway bulk package performance data that is equivalent to what they have for rail in the RSI-AAR TCAD. This view suggests that development of a highway bulk package database could lead to more favor- able consideration of highway transport, rather than less. Institutional Barrier Identification

110 challenges to cargo tank specifications and design. The 40-year experience of the RSI-AAR TCAD indicates that the accident data, analyses, and resulting actions taken based on industry consensus tend to provide reasonable means to effectively manage liability risks. Survey Feedback The industry organizations were asked to distribute a survey to their members regarding several aspects of a bulk package accident performance database. Survey respondents identified several concerns that would discourage participa- tion, as well as challenges to collecting requested information. Concerns Discouraging Participation Manufacturers, repair facilities, carriers, and shippers were asked to identify concerns that might limit participation in an accident performance data collection process. The principal concerns for all industry groups are confidentiality, liability concerns, and increases in the amount of paperwork required (see Table 67). Since the number of responses is considered a small sample size (less than 30), generalization of responses may include biases that cannot be detected. Measures Recommended to Increase Participation Survey respondents also suggested approaches to increase the likelihood that crash and tank damage information would be reported. These include the following: • Ensure the data are confidential or guarantee anonymity, at least from outside groups. • Review, redefine, and simplify existing reporting require- ments (this may include having law enforcement or Effect of Ranking One Type of Cargo Tank as Superior to Another Some stakeholders are concerned that an accident damage database may show that some types of cargo tanks are better than others. The database probably will demonstrate differ- ences among specifications; however, there is already qualita- tive understanding of this. An accident damage database would enable quantification of accident performance differences and better understanding of how specific design elements perform. Hazardous materials shippers and risk managers could use the resultant statistics to better inform their decision-making regarding cargo tank purchase and leasing decisions for prod- ucts with different hazard levels. It might also be possible to use data to support a reduction in carriers’ insurance premiums because their risks are better understood. An accident damage database would also provide a factual basis for discussions with regulators about whether proposed changes are justified and cost-effective. In general, an accident damage database could support a more accurate and refined approach to ongoing improvements in cargo tank design. Litigation Concerns Stakeholders have also expressed concern that the accident damage database may be used in litigation against them. This concern may be at the root of the substantial underreporting of accidents and incomplete reporting of accident informa- tion in PHMSA’s HMIRS. Several stakeholder representa- tives suggested that certain information provided in Form DOT F 5800.1 or in an accident damage database should be protected against disclosure under Freedom of Information Act (FOIA) requests and/or be restricted from use in liabil- ity cases. In general, equipment manufacturers, shippers, and carriers comply with federal regulatory requirements. Under these circumstances, federal preemption can be used to defend Table 67. Main concerns with providing bulk package accident performance data. Main Concerns Stakeholders Manufacturers Repair Facilities Carriers Shippers Confidentiality  Liability Concerns                Paperwork Cost Other Companies Not Participating Data Accuracy Time Requirements

111 a previous repair, the information identified as moderately difficult to collect includes the following: • Cause(s) of lading loss (i.e., shell puncture, bottom fitting damage, or rupture due to fire). • Location of the crack, gouge, puncture, or rupture from which the most hazardous material spilled. • Dimensions of the crack, gouge, puncture, or rupture from which the most hazardous material spilled. • Location of damage resulting in the loss of hazardous materials. • Shell or head thickness at damage locations resulting in the loss of hazardous materials. • Dimensions of damage at non-spill locations. • Shell or head thickness at non-spill locations. • Whether the crack or tear occurred because of damage to the fitting or appurtenance. • The type of wet line construction. The majority of the bulk package repair facilities indicated that the following information would be easy to collect: • Location of initial point of impact. • Dimensions of dent, crack, gouge, puncture, or rupture at an initial point of impact. • Shell or head thickness at an initial point of impact. • Shell or head thickness at the location from which the most hazardous material was spilled. • Dimensions of crack, gouge, puncture, or rupture from which the most hazardous material spilled. • Location of damage that does not result in a spill. • Location of damaged fitting. • Type of fitting damage. • Whether the cargo tank was repaired to specification, repaired to non-specification, or scrapped. • Tank shape. • Tank wall thickness. • Baffle and bulkhead location. • Type(s) of accident protection devices. • Type(s) of roll stability devices. • Presence of wet lines. In addition to the above information, some repair facilities also gather the following damage information: • Age of fittings. • Tank dimensions (out of round). • Frame, suspension, axle, and bumper damage. • Attachment damage. • Lighting damage. • Records of previous repairs and testing. • History of products hauled. emergency response officials communicate that the acci- dent meets reporting criteria). • Make reporting straightforward. • Automate the reporting system to accommodate wireless transmission of data. • Link multiple reporting systems to reduce redundant collection. • Create an incentive program such as linking reporting of accidents to a company’s safety rating, ensuring incident report records are up-to-date prior to issuing permits, or creating a tax incentive for compliance. • Require participation and impose greater consequences for non-compliance (e.g., a $5,000 fine with no mediation). • Have independent inspectors or assessors review/audit the data. • Launch a campaign to highlight the benefits of equip- ment improvements that will reduce the damage caused by crashes. • Provide useful feedback in a way that is easy for the driv- ers, mechanics, and others to understand (similar to the FMCSA cargo tank driver rollover prevention video). • Train police officers to note damage on accident reports. • Involve repair facilities in reporting. • Identify responsibility for the crash more clearly. Challenges of Data Collection Manufacturers. One manufacturer indicated that it col- lects accident data on crashes involving its bulk packages; however, the collection of these data is dependent upon whether they are aware of the incident and the incident type. This manufacturer also indicated that there is no set amount of information they collect; rather, information was only col- lected if it was deemed relevant. The other two manufacturers that responded do not collect data regarding crashes involv- ing their bulk packages. Repair Facilities. Repair facilities were asked a series of questions concerning the amount of training needed to ensure the collection of accurate information, the degree of difficulty anticipated in collecting various kinds of bulk package design information and accident damage information, and a rough estimate of the amount of time needed to collect this informa- tion. Of the seven repair facilities that responded to the survey, six indicated that some certification was necessary to ensure data quality. Three respondents indicated that the highest level of certification (Authorized Inspectors) was required. The damage information that was identified as most diffi- cult to collect is whether the damage occurred near a previ- ous repair and whether that repair influenced the structural integrity of the tank. Following the information concerning

112 ous materials are shipped using carrier-owned vehicles, the shipper requires the carrier to submit a full incident inves- tigation report including a root cause analysis. Therefore, in addition to variables recorded by various databases, as dis- cussed in Section 3, several shippers also report collecting the following: • On-board event recorder information (including speed and whether brakes were applied prior to the crash). • On-board drive camera (recording front and back of the vehicle before and immediately after the crash, including view of driver behavior at the time of the incident). • Pictures and witness information and interviews. • Information to determine whether an incident report should be filed (leakage amount, citation information, closing of public roadways, injuries, and environmental impact). Shippers also report several challenges in collecting accident information. These include the following: • Difficulty in gathering information from a carrier or carrier’s subcontractor. • Conflicting opinions with the police report regarding what caused the accident. • Inability of the driver to correctly recall events. • Lack of information concerning the events in the accident because not all vehicles/trailers are equipped with event recording devices. Identification of Institutional Barriers and Possible Solutions The success of an accident damage data collection system depends upon overcoming several institutional challenges. Based on the industry feedback discussed above, several of these institutional challenges were identified and are further considered in the following discussion. Possible solutions are also suggested. Institutional Barriers Associated with System Implementation and Operation Cost Barriers Barrier. During Fiscal Year 2011, the HMIRS cost approximately $2,255,000 ($1,945,000 annualized continu- ing resolution [P.L. 111-242 as amended] + $310,000 adjust- ments), and the research and analysis conducted by PHMSA cost approximately $435,000 ($423,000 annualized continu- ing resolution [P.L. 111-242 as amended] + $2,000 adjust- ments) (PHMSA 2011). These expenses would increase with The repair facilities were asked to estimate how long it would take to gather the above information. Four of the repair facilities estimated that it would require between 1 and 5 hours, while one company estimated that it would take 2 to 3 days. Another company indicated that estimating damage repairs “varies from tank to tank depending on whether the tank is ASME coded or not.” Estimates of the time required for collection of design information ranged from 15 minutes to 2 hours, with one repair facility estimating that 2 to 3 days would be required to collect the design data. Carriers. In addition to variables recorded by various existing databases, as discussed in Section 2, several carriers also report collecting: • Location information including GPS records and the number of miles from the home terminal. • Driver information including driver age, length of employ- ment, and years of experience. • Bulk package data including vehicle maintenance records. • Collision events data including electronic control module (ECM) or other on-board recorder information, photos of the accident scene and damage, root cause of crash, con- tributing factors, and accident reconstruction. • Package damage information including damage location, type, size, depth, wall thickness of damaged cargo tank, repair orders, adjuster investigation, and additional dam- age caused by righting equipment (in the case of a rollover). Since these data are being collected already, the additional burden of recording the data in a national database is mini- mal. However, the carriers also reported several challenges regarding the collection of crash and damage data. Working with local law enforcement at the accident scene and obtaining police reports in a timely manner was the challenge reported by the largest number of carriers. Other challenges to data collection include lack of software to collect specific data, lack of access to equipment due to impounding, increased response or remediation time due to inexperienced or insuf- ficiently trained emergency response teams, driver inability to take photos of accident (due to safety or law enforcement guidance), difficulty in preservation of incident scene, and lag time in getting a trained accident investigation engineer to the site. Some obstacles that need to be overcome in order for carriers to populate an accident damage database include the inability to describe and/or determine the condition of the vehicle if it is severely damaged, ability to collect the data in a timely fashion, and the threat of legal liability. Shippers. In general, shippers tend to collect informa- tion similar to that collected by carriers regarding hazardous material accidents. One shipper reported that when hazard-

113 dum database into regulation, a coordinated effort could be undertaken. This effort would focus on developing consen- sus among stakeholders regarding how to modify the form. Engaging stakeholder participation in developing an adden- dum to the highway bulk package Form DOT F 5800.1 would help mitigate issues that might otherwise hinder its adoption. However, developing consensus among the industry may be difficult because the industry includes a diverse group of car- riers and associations representing those carriers. A sufficient number of carriers must be convinced that such a database would be in their best interest. To achieve this critical mass, the individuals coordinating the revision effort should focus their initial effort on carriers already committed to risk-based approaches to safety. Barriers to Reporting Accident Data Several institutional barriers currently discourage indi- viduals from reporting accidents to PHMSA. These include drivers having to fill out paperwork without compensation, the possibility that legal repercussions may arise from the information provided to PHMSA, and the possibility that the individual required to report the incident is for some rea- son incapacitated by the accident and thus unable fulfill this requirement. These challenges will need to be overcome if the accident reporting system discussed in this report is to succeed in gathering sufficient and accurate accident infor- mation on which to conduct useful analyses. Reporting Responsibility Placed on Driver Barrier. PHMSA estimates that completion of Form DOT F 5800.1 by the individual reporting the accident would take 1.6 hours. This time estimate includes the following actions: reviewing the instructions, searching existing data sources for information, gathering the required data, and then completing and reviewing the report. The individual reporting the accident is a representative of the company in possession of a hazardous material during transportation. In the case of in-transit accidents, the driver of the bulk pack- age is involved in the reporting. Drivers are typically paid according to the number of hours or number of miles they have been driving, and paperwork is completed on personal time. Therefore, they may not be compensated for the time required to fill out a report. Any lack of compensation is a disincentive to the reporting individual and represents an institutional barrier to completing these reports. Possible Solution. When the driver involved in an inves- tigation is compensated for his/her time, this aspect of the institutional barrier is reduced or eliminated. the implementation of an addendum to Form DOT F 5800.1 and increased enforcement of participation. To justify this increase, the database must provide a return on investment. Since the addendum to Form DOT F 5800.1 to record high- way bulk package accident damage is expected to accrue acci- dent information at a relatively slow rate, this program may require several years of operation before the data can be used to assess accident performance and identify components or locations that would benefit from enhancements. Thus, a cost-effectiveness analysis should take this into account. Possible Solution. A possible method for justifying the expense of the addendum is to communicate the positive net return on investment for the RSI-AAR TCAD project. Since 1970, that database has been used to provide spon- sors with a scientific and statistical basis for evaluating the effectiveness and costs of new proposed rules and regula- tions. This approach has earned the respect of regulatory agencies and has resulted in more pragmatic, effective, and fact-based regulatory proposals. This database has been periodically evaluated for its effectiveness in aiding the industry to achieve various improvements in the safety of hazardous materials transportation by rail. Evaluations of the TCAD have consistently shown industry savings that were over 11 times the costs of implementing the database (T. T. Treichel to Phil Daum [principal investigator HMCRP Project 07], personal communication, 2/11/2011). The effort has led to safer, more secure, more competitive, and more profitable businesses. Another possible method to justify the database is to con- duct a pilot study with effective incentives for participation. The pilot study would enable realistic estimates of the additional cost of implementing an addendum to Form DOT F 5800.1. Additionally, the pilot study would provide further information regarding data accrual rates and lead to preliminary package performance estimates. These could be used to more accurately estimate the benefits of an accident database. Regulatory Barriers Barrier. Modifying the current Form DOT F 5800.1 to include the proposed fields would require a change to the current regulations governing compliance. Whatever changes are proposed must conform to the procedures outlined in the Administrative Procedure Act (APA). Therefore, this option would probably take several years and could involve opposition from the regulated community. The previous amendment to Form DOT F 5800.1 was revised many times over several years before being adopted. Possible Solution. To streamline the process of modify- ing the current Form DOT F 5800.1 or adopting an adden-

114 Reporting Responsibility of Carriers Barrier. Not only do companies lack incentive to report the incident, there is a disincentive because of the possible legal repercussions resulting from the information provided in the accident report. Possible Solution. Increased enforcement of existing requirements would motivate companies to ensure accidents are reported. An additional incentive may be provided by designing the system to rate a company’s participation in the program according to how consistently and comprehensively it reports accidents and tying these participation ratings into FMCSA’s safety rating system. However, if participation in the program is to be tied to the company’s safety rating, the accident rate for the carrier should not be generated from PHMSA-reported accidents. Rather, the accident rate for carriers should be generated from an external source, such as FMCSA’s database or a data set created from newspaper articles, and later matched to the carrier. For this incentive to be most meaningful and valuable, non-release accidents must also be reported. Privacy of Information/Legal Repercussions Barrier. Since the addendum would be an extension of Form DOT F 5800.1, similar to the information in the HMIRS, the additional data would be subjected to the cur- rent FOIA and, therefore, not protected. However, there is concern that the public availability of the stored data, under Option B, would have negative repercussions on the individu- als and companies responsible for reporting. The availability of these data for public use may either discourage individuals from reporting an accident or affect the accuracy and com- pleteness of data actually reported. Possible Solution. A possible solution would be the implementation of a “no-fault” provision so that information reported using the Form DOT F 5800.1 addendum cannot be used in liability cases. One way to implement a “no-fault” provision is to store information collected via the Form DOT F 5800.1 addendum separately from the Form DOT F 5800.1 information so that the damage data cannot be traced to the carrier, shipper, or other involved entity. Individuals respon- sible for reporting would then be provided with a unique receipt number as proof of compliance during company inspections and audits. In this way, the general public would be unable to correlate a bulk container damage report with a particular accident or bulk carrier. It is important to note that access to accident damage data is already available as part of the litigation process; therefore, collecting factual data through an addendum to Form DOT F 5800.1 would do no additional harm. Barriers to Obtaining Damage Information There are also several barriers to obtaining accurate damage information. These include challenges in recording crash and damage information at the crash scene and restricted access to the vehicle(s) and bulk package following the accident. Crash Scene Data Collection Barriers Barrier. The need to collect accident damage infor- mation often conflicts with the primary goal of emergency responders at the scene, which is to contain or prevent a spill and ensure the safety of individuals at the site. The presence of individuals at the crash site increases the risk of someone being injured. Furthermore, a goal of local law enforcement, to regain the use of the transportation route, conflicts with data collection because the incident scene may often be cleared prior to the information being recorded. Carriers reported several additional challenges that would result from trying to gather data while emergency responders and local law enforcement try to fulfill their responsibilities at the accident scene. Working with local law enforcement and obtaining police reports in a timely manner was the challenge most frequently reported by carriers. Other challenges to data collection include drivers’ inability to take photos of an acci- dent (due to safety or law enforcement guidance), difficulty in preservation of the scene, and lag time in getting a trained accident investigation engineer to the site. Possible Solution. Prior to the removal of the bulk pack- age and other vehicles from the accident scene, photos should be taken of the following: • Bulk package specification plate(s). • Hazard class placard. • Object(s) that damaged the bulk package in the accident. • Bulk package and any damage to the bulk package (these may be taken after the bulk package has been moved from its “resting” position). These photos may be taken by law enforcement personnel if a representative of the company is not present or otherwise unable to perform these tasks. These photos would be used to ascertain accident damage and other parameters if the bulk package was later unavailable. Additionally, an effort should be made to streamline the process of transferring photos and other information from local law enforcement to the carriers so that the information

115 can be reported. This may be accomplished by creating an industry website that has instructions and links to law enforcement request forms. Restricted Access to the Vehicle(s) and Bulk Package Involved in an Accident Barrier. Carriers reported two challenges to collect- ing information after the truck and bulk package have been removed from the accident scene: (1) the lack of software to collect specific data and (2) the lack of access to equipment due to impounding. These two challenges are typically over- come by estimating damage and accident conditions (i.e., the location and extent of damage and how fast the bulk package was traveling, respectively) thereby reducing the quality of data initially reported as part of the addendum. Possible Solution. The reporting system could allow for initial estimates to be revised once access to the vehicle(s) and bulk package has been granted. Conclusion The success of an accident damage data collection system depends upon overcoming several institutional challenges. These challenges and possible solutions to them include the following: • Cost barriers. Possible methods for overcoming these barriers include communicating the positive net return on investment for similar projects (such as the RSI-AAR TCAD) or conducting a pilot study with effective incen- tives for participation. • Regulatory barriers. Possible methods for overcoming these barriers include undertaking a coordinated effort that focuses on developing consensus among stakehold- ers regarding how to incorporate the accident damage data collection process. The initial attempt to gain support for such a regulatory revision could focus on carriers already committed to risk-based approaches to safety. • Reporting responsibility. Possible methods for over- coming these barriers include ensuring that drivers are compensated for time spent filing accident reports and increasing enforcement to motivate companies to ensure that accidents are reported. • Privacy of information/legal repercussions. Possible methods for overcoming these barriers include the imple- mentation of a “no-fault” provision so that collected information could not be traced to information currently collected by Form DOT F 5800.1. In such a scenario, car- riers would be required to provide proof of compliance during company inspections and audits. • Crash scene data collection challenges. A possible method for overcoming this barrier is to ensure that photos of the damage and accident scene are taken prior to removal of the bulk package from the accident scene. • Restricted access to the vehicle(s) and bulk package involved in an accident. A possible method for overcom- ing this barrier is to allow for initial estimates to be revised once access to the vehicles has been granted.