NTSB Board Meeting: Union Pacific Ethanol Train Derailment, Graettinger, IowaOn February 16, 2020 by Raul Dinwiddie
Welcome to the boardroom of the National Transportation Safety Board. I am Robert Sumwalt and I’m honored to serve as the chairman of the NTSB. Joining us today are my colleagues, Vice Chairman Bruce Landsberg, Member Earl Weener, Member Bella Dinh-Zarr and Member Jennifer Homendy. Today, we meet in open session, as required by the Government in the Sunshine Act, to consider the derailment of a Union Pacific Railroad train near Graettinger, Iowa on March 10th, 2017. The train was hauling 98 tank cars loaded with undenatured ethanol. 20 tank cars derailed, causing 14 of them to release about 322,000 gallons of ethanol, fueling a postaccident fire. Nobody died and nobody was hurt, fortunately, but there were three nearby homes that were evacuated and the accident caused $4 million in damage, not including the environmental remediation. NTSB investigators have worked hard to tease out the safety lessons of this derailment, pointing toward safer transportation in the future. Today, we will discuss the condition of the track in the area of the derailment known as the Estherville Subdivision. After the accident, Union Pacific undertook an impressive program of corrective actions. In just one example, crews replaced at least 8,600 crossties last year. Such ambitious catch-up work is noteworthy, it’s necessary and it’s appreciated. But rail infrastructure generally deteriorates over time, not overnight. And as they say, no good deed goes unpunished. The extent of postaccident actions, while welcome, also hints at the inadequacy of UP’s pre-accident maintenance and inspection program. The railroad is supposed to look for and fix unsafe conditions as a matter of course. The Federal Railroad Administration oversight is supposed to ensure that the railroad at least maintains the tracks to meet the minimum Track Safety Standards specified in the regulations. Today, we will discuss the inadequacy of UP’s track maintenance and inspection program, and the FRA’s insufficient oversight. The derailment also provided us with an opportunity to revisit fiery releases of ethanol from breached DOT-111 tank cars. These tank cars must be phased out of ethanol service by mid-2023. One thing we examined was the industry’s progress in meeting this deadline. As we will discuss, if this unit train consisted of new DOT-117 tank cars, the tougher design would have mitigated or prevented product loss due to tank shell breaches, head breaches and damaged top fittings. But the reality is that most tank cars used in ethanol transportation are still the older legacy DOT-111 tank cars. And although retrofitting or replacing with DOT-117 cars is required by 2023, the Pipeline and Hazardous Materials Safety Administration has created no intermediate milestone schedules. This stage is set for the possibility of a missed deadline. Just as we saw with the congressionally mandated Positive Train Control, there was a deadline and as we approach the deadline, the industry raised their hands and said we can’t meet it. And I think, we, here at the agency are concerned that the shippers are going to get down close to that 2023 deadline and say oh gosh, we can’t meet that deadline. Finally, this derailment brought to light one more fascinating wrinkle, the behavior of undenatured ethanol in this release. A denaturant is an agent added to ethanol to make it unappealing, harmful or potentially lethal to drink. In practical terms, once the ethanol is no longer drinkable, it can be offered as fuel without being subjected to beverage tax requirements. Now, denatured ethanol is not like decaffeinated coffee. Ethanol doesn’t start out with denaturants that you need to distill out. Instead, you need to add something to denature ethanol, usually something harmful or lethal if ingested. Denaturing ethanol is, quite simply, a policy-driven process that separates ethanol for beverages from ethanol for fuel or other industrial uses. And I did a little digging and turned up this bit of historical perspective. According to Time Magazine, in 1927, the government changed the formula for denatured ethanol. And the reason, think about it, prohibition. In prohibition, people had begun drinking, had begun drinking household denatured ethanol. The previous formula was simply not harmful enough to deter people from drinking it. So we made it twice as deadly. Then as now, the purpose of denaturing was to create different categories of ethanol for different uses. Back during prohibition, one category was an illicit beverage. Only by poisoning the illicit beverage could one render it safe, legal for sale. Today, we denature ethanol to create fuel ethanol which therefore is not taxed as a beverage. In fact, often we add 2.5% of one kind of gasoline to ethanol to make it suitable under the law for mixing with another kind of gasoline. None of this, of course, falls within the NTSB’s purview until denaturing affects transportation safety. And in the Graettinger accident, the ethanol released was undenatured and this seems to have had a beneficial effect on safety. Investigators found milder damage in this accident than in previous accidents with the same type of tank cars, but those involved denatured ethanol. So we will consider whether it might make more sense to leave ethanol undenatured prior to shipment, from a transportation safety point of view. Today, the NTSB staff will briefly present the most pertinent facts and analysis found in the draft report. Our public docket available at www.ntsb.gov contains over 1,200 pages of additional information, including photos, interviews and inspection records. Staff has pursued all avenues in proposing findings, a probable cause and recommendations to the board. And the order of today’s board meeting will be that staff will present topic-by-topic summary of the investigations. We on the board will then question staff. We will propose and vote on amendments as necessary to ensure that the report as we adopted today truly provides the best opportunity to enhance safety. Now, Managing Director Dennis Jones, if you’ll would kindly introduce the staff. Good morning. Good morning, sir. Thank you, Mr. Chairman. A few announcements before we begin. I kindly request the folks here in the boardroom if you have not already done so, please silence your mobile phones and other electronic devices. There are two exits in front of the auditorium, one to my left, one to my right, go down the stairs out the door and follow the illuminated exit signs of the part of facility. You may also exit to the rear of the auditorium and proceed out the glass doors you entered, go up the stairs and exit straight ahead through the large glass doors to the outside. In the event of emergency, please walk quickly to the nearest exit and make your way to the outside following the instructions of NTSB staff personnel. Do not delay. Do not return to the boardroom until instructed or advised to do so. If you have any questions or concerns, please don’t hesitate to contact any NTSB personnel. Seated at the panel, unless otherwise noted, are staff members of the office of Railroads, Pipelines, and Hazardous Materials. Starting on the front row to my right is Robert Hall is the office director. Next is Mike Hiller who’s the investigator in charge. Next is Jim Southworth who handles mechanical aspects. Next is Joe Gordon who will be handling track aspects. Next is Dr. Erik Mueller who will handle metallurgy aspects and he is from the Office of Research and Engineering also known as RE. Behind him on the second row is Paul Stancil who handle hazardous material aspects. Next is Kevin Renze who handle vehicle performance aspects and he’s from the office of RE. Next is Dave Bucher who’s the division chief for railroads and he will be handling operations. Next is Barbara Czech, the deputy director of the office of RE. next is Kathy Silbaugh who’s the general counsel. Next is Paul Sledzik who’s the director of the Office of Safety Recommendations and Communications also known as SRC. Behind him on the third row is Troy Lloyd who will handle visuals and he’s also the timer. Next is Gena Evans who’s the editor and next is Mike Hamilton who will handle safety recommendations and he’s from the office of SRC. We will start the presentation with an overview of the accident by the investigator in charge, Mike Hiller. Thank you, managing director, good morning. On March 10th, 2017 at 12:50 a.m. Central Standard Time, eastbound Union Pacific Railroad ethanol train, UEGKOT-09, with three locomotives, 98 loaded tank cars and two buffer cars derailed near Graettinger, Iowa. The train derailed when the rail broke under the train between the lead locomotive in the 20th car prior to the west approach of the Jack Creek Bridge. 20 tank cars in positions, 21 through 40 derailed. 14 of the tank cars released about 322,000 gallons of under natured ethanol fueling a postaccident fire. There were no injuries and no mandatory evacuations. About 400 feet of track and a railroad bridge were destroyed. The lead locomotives event recorder data showed that the train was traveling about 28 miles per hour when there was an emergency application of the air brakes without input from the train crew. The lead locomotives and 20 cars crossed the bridge without derailing. According to the event recorder, the emergency brake activated at about 12:50 a.m. The three-dimensional animation you were about to see shows the estimated derailed car pileup sequence. NTSB’s Office of Research and Engineering reconstructed the train motion to help assess the initial energy state, path and mechanical impact damage of each derailed car. On rail vehicle motion was reconstructed from locomotive event recorder external video and external audio data as well as car, track and wayside infrastructure data. Derailed car motion was derived from car on rail in post-position and orientation data as well as the entry speed of the trailing cars recorded by the rear locomotive. Interpolation was used to define the path of each derailed car between its on rail and at rest positions. The resulting derailed car motion is not strictly time accurate, kinematically consistent or the product of a fully validated first principles based physics model. The animation will show two views of the estimated derailed car pileup sequence. The local time after midnight, lead and trailing locomotive speed and throttle position are presented in the right hand side upper frame. The lead locomotive external camera view forward is shown in the lower frame. Lead locomotive external audio data you will hear has been synchronized with the train motion and external video. The uncommanded emergency brake application, ethanol ignition and combustion events and postaccident drone survey data will be annotated. This animation does not attempt to portray the actual time of day, season, weather or lighting conditions at the time of the derailment except as recorded by the lead locomotive external camera. Car collision dynamics, deformation, component separation, damaged track and bridge damage and environmental damage are not depicted except as documented by the aerial drone survey. Please play the animation. (train hisses) (train honking) (train honking) (train hisses) (train honking) (train honking) The tank cars involved in this accident were built to DOT-111 specifications. 10 of the derailed tank cars breached from mechanical damage. The NTSB has identified DOT-111 tank cars as having a high probability of releasing hazardous materials in accidents and has recommended crashworthiness requirements to improve safety. The Pipeline and Hazardous Materials Safety Administration responded with new federal regulations for tank cars used in flammable material service accordingly. Tank cars used to transport ethanol must be constructed to new DOT-117 specifications and existing tank cars must be removed from ethanol service by May 1st, 2023. Mr. Stancil will further discuss this in his presentation. As part of this investigation, NTSB’s Office of Research and Engineering studied the strength of 90-pound rail, rail which was the type involved in this accident. Compared to some of the heavier types of rails used in the railroad, 90-pound rail is light and this investigation sought to understand if this rail is adequate for heavier, hazardous material tank car loads rolling across the rails today. Dr. Mueller will discuss the results in his presentation. The on-scene investigation team. NTSB staff supporting this investigation. The parties to this investigation. The issues discussed in staff presentations will cover the adequacy of the Union Pacific’s track maintenance and inspection program, the adequacy of the Federal Railroad Administration’s track inspection oversight and the transportation of fuel ethanol without the use of volatile organic chemical denaturants. This concludes my opening presentation. Mr. Southworth will discuss the mechanical details of this accident. Morning. The unit ethanol train was traveling in an easterly direction. The right side wheels of the locomotives and railroad cars were running on the south rail. Arrows indicate the direction of travel and the location of the south rail. On March 9th, 2017, at the Green Plains refinery in Superior, Iowa, the train crew, about 2 1/2 hours prior to the derailment, the train crew performed a Class 1 brake test and initial terminal inspection. Green Plains is less than 20 miles from the accident location. Crews are required to inspect and conduct a brake test on all cars in the train at the location where it’s originally assembled to verify that the brake system on all cars in the train are operating as intended prior to departure. The crew found no problems during the inspection and brake testing. The train separated between the 20th and 21st cars. Equipment that had not derailed was moved to the east and west away from the derailment area. Postaccident inspections and air brake tests of the non-derailed cars and the train revealed very few defects, none of the defects observed on the cars were causal to the derailment. All cars inspected and tested and revealed good integrity of the train brake system. On March 11th, 2017, investigators also examined the derailed tank cars, truck assemblies and wheel sets during the postaccident recovery. All wheels were found to be intact except for two wheels that were fractured due to derailment forces. Several wheels showed indications of thermal exposure consistent with the tank car fires. Brake equipment was also inspected to the extent possible. No preexisting conditions were found that would contribute to the cause of the derailment. Several south side wheels on 14 of the first 20 cars in the train exhibited fresh horizontal marks on the wheel treads. These two photographs show the location and position of horizontal marks on wheels that travel over a discontinuity in the south rail. The yellow arrows point to the single marks on the wheel treads. Investigators did not observe any marks on wheels that traveled over the north rail. None of the lead locomotives, the buffer car and the non-derailed cars at the west end of the train, the 41st through the 98th car and the distributive power unit exhibited any horizontal wheel marks on wheel treads. This is another photograph of wheel tread damage to the wheel of the 11th car. These pictures show south side wheels of the 20th car from the head end of the train. On this car, all four wheels that rolled over the south side rail showed marks on the tread. Notice in the top right photograph that there are two arrows pointing to two marks on the wheel. The arrows in the two bottom photographs show a closer view of marks on wheel treads. In summary, the marks on the south side wheels are indicative to a rail break under the train behind the lead locomotives. The evidence showed it began to appear on the fourth car behind the lead locomotives. The damage progressively worsened as more cars traveled over the break until it catastrophically failed and derailed the train. This concludes my presentation. Joe Gordon will now discuss track and engineering. Good morning. I will discuss the Union Pacific’s maintenance and inspection of the track structure of the Estherville Subdivision in Graettinger, Iowa and the Federal Railroad Administration’s oversight of Track Safety Standards on the accident subdivision. The Union Pacific Estherville Subdivision is a single main track between milepost zero and milepost 79.3 with one passing siding. The subdivision’s average daily train count was one train every other day. The track in the accident area was made up of 90-pound continuous welded rail fastened through top plates to standard eight-foot wooden crossties using conventional track spikes. At the time of the derailment, UP operated freight trains at a maximum operating speed of 30 miles per hour or lower and UP does not operate any passenger trains on the Estherville Subdivision. In response to a June 3rd, 2016 UP high-hazard flammable train derailment in which the content of three tank cars fueled a bridge, fueled a fire at a bridge ramp at the approach to Interstate 84 in Mosier, Oregon, UP entered into a compliance agreement with the FRA on December 22nd, 2016. This compliance agreement required UP to agree to remedial action to improve compliance with FRA minimum Track Safety Standards. Because of the agreement, UP was to conduct required high rail or walking track inspections of the Estherville Subdivision twice weekly as well as additional automated track geometry testing and ultrasonic rail testing. Investigators reviewed UP track inspection records from July 1st, 2016 to March 9th, 2017. The records revealed that the tracks were being inspected at the required frequency. UP inspectors documented 49 locations between milepost 44.5 and milepost 78.4 with defective tie conditions between July 1st, 2016 and November 4th, 2016. Records also showed that UP track inspectors had documented marginal tie conditions on the subdivision meaning that the ties were at or near the end of their service life. Investigators reviewed FRA track inspection records from March 2015 through the day of the accident. These are the findings related to the reporting of crosstie conditions. In March 2015, one FRA inspection report contained a comment regarding marginal crosstie condition approaching defective conditions. Again, in July 2015, an FRA inspector provided comment on an inspection report about the marginal crosstie condition for the class of track being operated in the area. In addition, the inspector said that those crossties were close to the end of their life and another winter cycle would most likely put those areas in non-compliance. The inspector also emphasized that this was an ethanol route. On August 10th, 2016 FRA inspectors documented 10 locations with crossed eyes that did not meet the minimum requirements for FRA Track Safety Standards. The inspectors characterized the crosstie condition as being marginal and in many areas, defective, emphasizing in the report that this was an ethanol route and the risk was evident. Although FRA track inspectors identified some defective crosstie conditions, not all such defects were recorded. The record showed an FRA inspector last examined the track in the area of the derailment December 13th, 2016 noting no defective conditions, however, the report contained a comment stating that the track had marginal tie condition between Superior and Emmetsburg which is included in the derailment location. No civil penalties were recommended in any of these reports. Investigators conducted walking inspections of the derailment area, during these inspections, found seven defects that did not meet the FRA minimum tracks standards. They included four crosstie distribution defects and one insufficient number of crossties in a rail segment. In May of 2017, when the FRA returned to the Estherville Subdivision and conducted walking and high rail inspections, they documented 78 defective track conditions. Of those, 51 were locations with defective crossties and five were locations with rail fasteners not maintaining track gauge. A clear demonstration that all defective crosstie conditions were not being reported prior to this accident. No civil penalties were recommended in any of these inspection reports. As discussed in the report, crossties do not deteriorate in a matter of days to the extent that they are split broken and unable to hold fasteners but do so over time. In summary, staff found that UP did not maintain the Estherville Subdivision in accordance with FRA minimum Track Safety Standards or UP’s internal track maintenance standards. UP supervisors and managers were not ensuring defective crosstie conditions were being identified, reported and remediated in accordance with UP’s track maintenance standards or FRA Track Safety Standards. All defective crosstie conditions on the UP Estherville Subdivision were not identified by FRA in the two years prior to the derailment and all available FRA enforcement options such as a recommendation for civil penalties were not used. Staff has proposed recommendations to address these areas. This concludes my presentation. Next, Dr. Erik Mueller will discuss metallurgy and finite element modeling. Good morning. The materials laboratory portion of the derailment investigation involves several aspects which I will cover in this presentation. First, I will briefly touch on the on scene rail recovery. I will discuss the testing performed on selected pieces of rail sent to the materials laboratory including results pointing to the likely point of derailment on the track. Then I will discuss the finite element modeling performed by the vehicle performance division to evaluate the use of 90-pound rail with today’s heavy freight loads. Staff recovered about 390 feet of north rail and 385 feet of south rail while on scene. This left about 10 feet of the north rail and 15 feet of the south rail unrecovered. This figure shows a graphical representation of the rail reconstruction with the rail fragments labeled from zero to 15 from west to east for the purposes of this investigation. Several rail pieces exhibiting prominent features typical of damage incurred during heavy freight rail use were sent to the materials laboratory for further examination. This included nine pieces of north rail and five pieces of south rail. The materials laboratory performed an extensive battery of tests on the rail pieces received including non-destructive examination, scanning electron microscopy, optical metallography, hardness testing and chemical composition testing. One of the rail pieces the materials lab focused on was a fragment labeled 10S located on the Jack Creek Bridge 4.6 feet east of the transition of the track from ballast to the west side of the bridge. Just west of this rail piece was the location of approximately 12.6 feet of south rail that was not recovered during the on-scene investigation. This figure shows the aforementioned rail fragment viewed from the field side corresponding with the train having moved from left to right towards the east direction. Of note was the west facing or left fracture surface in the figure. The fracture service exhibited had batter, a type of longitudinal impact damage consistent with the exposed surface having been repeatedly impacted by passing wheels. No other rail fragment that was recovered exhibited this degree of head batter on a fracture surface. The gauge side or side opposite of the field side of the rail exhibited features such as head checking and subsurface shell cracks consistent with rolling contact fatigue. Surface and subsurface cracking consistent with head checking and shelling rail head defects were all present on section 10S. These defects have been known to lead to fatigue cracking and premature rail fracture in previous train derailments. While these features were present in this and other pieces of rail, none of the fractured surfaces exhibited evidence that these defects had developed to the point that they would cause premature fracture of the rail. Moreover, microscopic examination of the west fracture face revealed cleavage facets consistent with low ductility overstress fracture. In addition, the heads of the recovered pieces of rail were inspected using ultrasonic testing and non unrevealed crack indications consistent with internal cracking. However, the degree of the head wear and damage on the other pieces of rail was generally inconsistent with the widespread pre-existing damage of section 10S. The fact that the 10S piece exhibited the most subsurface damage, extensive head batter and was located closest to the transition from the ballast to the timber bridge where the 12.6 feet of missing rail would have been points to this rail piece as being the most likely point of derailment. A computer simulation tool called Finite Element Modeling was used to determine if the 90-pound rail from the accident who withstand the deformation and stresses under tank car wheel loads. Two three-dimensional models of the 90-pound rail were constructed. The best case model based on drawings provided from Union Pacific modeled an unworn standard 90-pound rail profile with stiff track support conditions. The worst case model seen on the right in the red on the figure used a worn rail profile based on measurements collected on accident rail profiles and use less stiff degraded track support conditions. The gray area on the right profile shows the amount of worn head area relative to the new rail profile. This figure shows a graphical representation of the results with red representing the highest tensile stresses, blue representing the highest compressive stresses and green being neutral or near zero stresses. With typical tank car wheel loads applied, the model outputs showed that the worn rail profile seen on the right exhibited higher transverse or left to right stresses in the head region. This was due to the reduced head area from the wear and degraded track support conditions which supports the loads of the rolling train wheels. The peak transverse stress could be about 50% higher for the case with the worn profile and less stiff support. Excessive transverse stress in the head region has been known to create cracks in the vertical direction known as vertical split head failures. While the worn track with the degraded support exhibited highest levels under wheel loads compared with the best case condition, the increased stresses were not sufficient by themselves to lead to track failure. The finite element model showed that assuming proper maintenance, 90-pound rail can withstand the wheel loads of fully loaded tank cars. In summary, staff examined the fracture services of the rail recovered from the accident, all of which exhibited features that were consistent with overstress fracture. Many of the rail pieces exhibited pre-existing defects that are known to lead to premature fracture and much of the underlying track support had degraded. Based on the results of the finite element modeling study, staff concluded that a deteriorated track structure such as worn rail and degraded track support conditions would cause more track movement and higher stresses in the rail. In particular, the study demonstrated that worn 90-pound rail on degraded track support result in higher transverse stress in the rail head region exposing the rail to increase risks of failure due to a vertical split head, however, the study also showed no indications that the track was in danger of failure by overstress assuming that the loads, track and support conditions in the model were representative of actual conditions. While none of these individual factors by themselves caused the derailment, their combined effects and the preponderance of evidence indicates that the likely derailment location was a section of south rail near the bridge which exhibited the most microscopic and macroscopic damage. Therefore, based on the observation of the fresh horizontal impact damage observed on the wheel tread of the fourth through 20th non-derailed cars, examination of the rail recovered from the accident and the condition of the cross tie structure on the Estherville Subdivision, staff concluded that the train likely derailed from a broken south rail that occurred prior to or at the 20th car as it was traveling over the west approach of the Jack Creek Bridge resulting from UP’s inadequate track maintenance and inspection program and FRA’s inadequate oversight of the application of Federal Track Safety Standards. This concludes my presentation. Mr. Paul Stancil has the next presentation focusing on the hazardous materials issues in the investigation. Good morning. My presentation will discuss the possible safety benefits of shipping ethanol without the addition of volatile chemical denaturants and the progress of replacing DOT-111 tank cars used to transport ethanol to achieve compliance dates for better performing tank cars as mandated by the FAST Act. The EPA has established incremental increases in the amount of ethanol required to be blended into the US fuel supply with a target of 36 billion gallons annually by 2022. Because railroads account for about 70% of ethanol transportation from biorefineries to fuel blending facilities, ethanol is currently ranked the top hazardous material transported by rail. Alcohol and Tobacco Tax and Trade Bureau beverage tax regulations disincentivize shipment of under natured ethanol and require the use of volatile chemical denaturants, typically, natural gasoline to render the spirits from alcohol fuel plants unfit for beverage use. Otherwise, distilled spirits are taxed at a higher rate if the ethanol is diverted to beverage use. The use of denaturants substantially alters ethanol’s physical and chemical properties and also introduces toxic compounds. And under the Environmental Protection Agency, Renewable Fuel Standard provisions, ethanol does not become a renewable fuel and assigned renewable information numbers or RINs until it is first denatured in accordance with TTB regulations. These RINs are part of a complex system for tracking compliance with the renewable fuels program and are bought, sold and traded amongst obligated parties. While fuel ethanol produced for domestic use in the US must be denatured, there has been a recent substantial increase in the number of undenatured fuel ethanol shipments for export to foreign markets. According to the Renewable Fuels Association, some foreign consignees prefer to receive their ethanol in its undenatured form and typically do not denature ethanol. This graph shows a decline in denatured fuel ethanol exports while in 2017, the amount of exported undenatured fuel ethanol almost tripled over previous years. Undenatured ethanol is shipped from the same production facilities that produce denatured ethanol using the same types of rail cars in the same transportation routes. The accident in Graettinger, Iowa is the first and only derailment with a hazardous materials release and fire involving a unit train of tank cars transporting fuel ethanol in its undenatured form. Investigators noticed that the milder postaccident fire damage did not compare with the outcomes observed from previous denatured fuel ethanol accidents, several of which produced energetic fireballs, rocketing car parts and significant fire damage resulting in large thermally induced shell tears. Yet, the tank cars involved in the Graettinger accident were the least robust legacy DOT-111 tank cars. Furthermore, the presence of toxic denaturing constituents has caused significant environmental contamination and has greatly complicated site remediation efforts following previous accidents with denatured ethanol but because the undenatured ethanol involved in the Graettinger derailment did not contain these toxic compounds, the environmental consequences of this accident were much less severe than observed in previous accidents. The need to denature 23 billion gallons of fuel ethanol currently transported annually for domestic use with toxic volatile organic compounds merely to discourage its uses of beverage potentially poses an unnecessary risk of death, injury and environmental contamination when released in transportation accidents. While there is limited accident data to compare the reduced thermal damage to tank cars, the lack of energetic fireball eruptions and less environmental impact in the Graettinger accident, the similar railroad accidents involving denatured ethanol, a safety benefit may be derived from transporting ethanol without the use of denaturing chemicals. However, government and industry stakeholders have not studied what effect the absence of denataruing might have on the safety of transporting ethanol and unit trains and therefore, staff has proposed a safety recommendation to address this issue. Mechanical damage to the legacy DOT-111 tank cars involved in this accident was typical in scale and breaching mechanism to that scene and many other accidents the NTSB has investigated. As mentioned earlier, the derailed tank cars sustained mild thermal damage. 322,000 gallons of ethanol was released from 14 of the 20 derailed cars. Of these, 10 tank cars sustained breaching impacts to heads and, or shells. One tank car exhibited a very small thermal tear, three tank cars only reduced released product because unprotected top fittings or bottom outlet valves were damaged. A large number of breach tank cars is further evidence of the DOT-111’s lading-retention performance. Current federal regulations responding to the previous NTSB safety recommendations call for their removal from ethanol service by May 1st, 2023. Since October 1st, 2015, federal regulations have required better tank cars for use in high hazard flammable trains such as the DOT-117 equipped with impact resisting head shields, thicker shells, thermal protection systems, top fitting protection and bottom outlet valve protection. Nevertheless, the train, the accident train was composed entirely of legacy DOT-111 tank cars with the least crash resistant feature is available for flammable liquid service. The Association of American Railroads reports that as of the first half of 2018, about 1/3 of the tank cars shipping ethanol today are now DOT-117 compliant. There remains about 18,700 DOT-111s and over 3,000 CPC-1232 tank cars that must be replaced or converted to DOT-117 specifications. According to the AAR, this would require about 350 tank cars per month to be replaced in order to achieve the 2023 deadline for replacing legacy DOT-111 and non-jacketed CPC-1232s and the 2025 deadline for replacing jacketed CPC-1232 tank cars and ethanol service. However, staff is concerned about the potential for critical shortages of DOT-117 tank cars for flammable liquids transportation if tank car owners fail to take adequate measures to ensure the timely completion of their retrofits and replacements. Prior to publication of these new PHMSA regulations, the NTSB expressed concern about the lack of intermediate progress requirements for completing modifications or replacements and the lack of transparency to hold the rail car owners accountable for progress toward implementing safer tank car fleets. As a result of the 2015 accident of a unit train of CPC-1232 tank cars hauling crude oil in Mount Carbon, West Virginia, the NTSB called for an aggressive implementation schedule with intermediate milestones and a publicly available reporting mechanism. Safety recommendation R-15-16 is currently classified Open-Unacceptable Response because PHMSA has not established a clear set of intermediate metrics that it could use to evaluate the safety improvement progress. During the first two years following the issuance of regulations requiring better tank cars for high hazard flammable trains, ethanol fleet owners had made very little progress updating or replacing their DOT-111 tank cars. Since most shippers are not rail car fleet owners and generally leased their cars, they must rely on third-party leasing companies to ensure the availability of replacement tank cars in advance of regulatory deadlines. At the time of this accident, four of the seven rail car leasing companies that supply tank cars for the Graettinger accident train had predominantly DOT-111 and CPC-1232 tank cars in their fleets. Three of the leasing companies had no DOT-117 tank cars at all. While recent AAR ethanol fleet statistics show remarkable progress was made during the first half of 2018, staff believes that without mandated progress requirements, achievement of the deadlines may be overly reliant on future market and economic conditions. Staff has proposed to reiterate an existing safety recommendation to address this issue. This concludes staff presentations. Mr. Hall. Thank you. Chairman Sumwalt, Vice Chairman Landsberg and members of the board, this investigation is about three things. First, the track and the track structure. The track structure was neglected and the inspections that looked at the track structure failed to identify needed repairs. We’ve seen that before. We saw it in Metro-North, we saw it in WMATA. What needs to be done is known by the industry, it just needs to be done. The second item was the replacement of the DOT-111 tank cars. Staff is concerned about replacing them in time. As we’ve seen, recent efforts by the federal government to actually increase the use of ethanol and we’ve also seen in the past where important safety mandates made by the Congress have not been met as in the case of PTC where at the last minute, a extension was required. Lastly, is the safety of the transport of the hazardous material ethanol. In my career, 38 years since I graduated from college, I have worked in the production, transportation and use of hazardous materials including some of the most hazardous materials that we use in our society. High level nuclear waste, hypergolic rocket fuels, poisonous chlorine, all of these fuels I’ve worked with in the past. Never before have I seen a case where we actually take steps to make a hazardous material more hazardous before we ship it and it just doesn’t make sense. There are safety benefits that can be achieved and we should look to get those safety benefits. Thank you and the staff would appreciate your questions at this time. Mr. Hall and to the entire staff, thank you for, first of all, very good investigation and for very good presentations and Mr. Hall, thank you for wrapping it up with those three points. We will turn now to the board members for questioning beginning with Vice Chairman Bruce Landsberg. Thank you, Mr. Chairman and thank you, staff. Really truly an excellent report. I did a little research as well, Mr. Chairman. Do you know when Prohibition ended? Well, it was 1933 when Prohibition ended and it seems like, there seems to be an undue concern regarding people drinking ethanol as it were. I’m gonna use the terms pure versus denatured because undenatured seems like too much of a tongue tie. What’s the difference, Mr. Stancil, in a flash point of pure alcohol versus the denatured variety? So the flash point of pure ethanol, 100%, on denatured ethanol is roughly 55 degrees Fahrenheit. The undenatured ethanol, the flashpoint may vary depending on the percentage and the formulation used. We’ve seen it reported as low as minus five degrees Fahrenheit to as high as about 19.4 degrees Fahrenheit. So would it be fair to say that it is far more potentially explosive when it’s denatured? Well, the use of the term explosive may be a little troubling there. There are a number of parameters that can be used to determine the hazardous nature of the two materials and frankly, the FRA has done some research comparing ethanol and crude oil, its behavior and the differences between the two. However, this research has looked at ethanol as denatured ethanol but has used the thermodynamic properties of undenatured ethanol because they’re more available. It’s basically not thoroughly understood and I think that’s one reason why we’re asking for this study. Understand. I guess one of the big driving points here is the issue of taxation and so forth. Was there any review or study about how this particular train which was, I guess, destined for foreign export carrying pure alcohol, was it taxed in any way? And you may not know the answer to that but it seems like we have sort of a false equivalency here of taxation versus non-taxation and the threat to the environment and the nature of the hazardous material. I’m not formulating this very well but it seems like we’re getting down to economics as far as the decision being made as to how this is shipped. Yes, so, my understanding there there is an excise tax on drinkable spirits. It’s, I believe, $13.50 per gallon at this point. In this case where the material was shipped as undenatured, there is, my understanding is there’s some sort of a bonding requirement to ensure that the material arrives, the amount of material that’s shipped equals the amount of material that arrives for export. Do we have any data whatsoever to show the amount of pilfering that takes place from tank cars? Well, we did contact the UP and asked whether there had been any previous incidents of that and they reported to us that there had been none that they’re aware of. Moving to track, are there different rules that exist to differentiate track inspection and requirements where hazardous shipments are being carried? Not under the current regulation. That did come into play in the compliance agreement that compliance agreement was targeted to how hazardous train routes but in the minimum standards, there are no differentiation between the tracks that handle the high hazardous flammable material trains. Would that seem like it might be a good idea? So, I realize we have a lot of different track and some tracks will carry hazardous materials, others won’t and would it make any sense to have a differentiation between hazardous tracks versus non hazardous tracks? Yes, and that has been something that FRA has been working with since the shipment of ethanol and crude oil started moving by rail more regular. They’re looking at risk assessment, risk model as far as their minimum standards. Very good. Thank you, Mr. Chairman. You’re quite welcome, Vice Chairman. Member Weener. Thank you. I’d like to talk, question about failure modes of the track. What are the predominant failure modes of the track? I’m assuming the track at this point means the rail itself as well as the ties and ballast and so forth. So looking at the statistics from this year, they’ll go from January 1 through, there’s a little lag but through the end of July, the leading cause of main track derailments is broken rails. When you when you factor in all of the different types of broken rails, where the crack originates, there are many different ways that rail fails. So if you add all those together, the broken rail is the leading cause of main track derailment. It’s actually all tracks, if you look at all tracks, broken rails, the leading calls there as well. The second which, when you factor in all broken rail or the top five causes of broken rails that would be over 10% of all derailments would be broken rails and the next track calls would be 4.6% which would be what wide gauge where the distance between the rails is greater than allowable. So some of these are immediate failures and some of them are latent failures. Is that an accurate statement? That’s correct. What are the precursors for real failure. So how can they be detected? So rail failures, there are a lot of work that goes into better detection of internal rail defects. That’s one of the, with it being the high nail in track calls derailments. The industry is definitely looking at ways to improve identifying internal rail defects and currently, there’s work going with continuous testing to where now, the standard way that the tests are conducted is the vehicle sits on the rail with the ultrasonic testing equipment, they test at a low speed, normally, around 10 miles per hour and then when they identify a defect, they have to stop and back up and hand test that defect to actually classify the size of the defect. So it’s tedious work. Continuous testing would allow the railroads to test from one end. If there’s a hundred mile territory, they would be able to test from one end of that territory to the other just at a continuous speed and then analyze the data later but that’s one of the biggest improvements that they’re looking at right now. The predecessor to that was actually walking the track, is that correct? So, these ultrasonic rail tests are in addition to walking the track. The carrier’s still required to conduct weekly or twice a week visual inspections of the track and these can be done either by traversing the track in a high rail vehicle or walking the track. Those still have to be done so there’s still a visual inspection of the rail, however, most of these rail failures occur from an internal rail defect that’s undetectable with your eye. So it has to be some sort of non-destructive testing to actually identify what the problem is? Yes. Now, this rail in the accident area was 90 years old which is just a little after the end of World War I. Is there an aging process that occurs with rail? It’s more, more concern as the rail wear from the movement over the rail as well as the processes for manufacturing rail have improved greatly. Have the materials also improved, the alloys used? Yes, the alloys as well as the manufacturing process for modern rail have improved considerably but the rail that is, as always, I say, there’s nothing inherently wrong with those compositions or their manufacturing processes in and of themselves. It’s again, more, as Joe is saying, the wear to the head of the rail over time. Very good, thank you. Thank you, Member Weener, and Member Dinh-Zarr. Thank you, Chairman and thank you for launching to this accident. I know ’cause–
It was cold. A cold launch, yes. I was very impressed with this report. Thank you to staff. The animation with the synchronized sound was very helpful, I thought, and I think this is an excellent example of how careful analysis of one case which we often have to do can unearth unexpected safety concerns or in this case, a safety benefit of a undenatured ethanol or pure ethanol. So, I think the vice chairman mentioned and made his points very well and clarified this, that we’re actually adding, and Mr. Hall, that we’re actually adding toxins to a material that makes it unsafe for both from a transportation and environmental perspective. But I know that, Mr. Stancil, you mentioned exporting ethanol and I’m wondering do other countries also denature ethanol? Yes, there is a certain amount of denatured ethanol that is shipped from the US for export but for the most part, it’s undenatured. And are there any countries who do it there or do we always do it here and then we either denature it or not and then export it? I really don’t know the answer to that question. The requirement that do it here though is sort of tied into the EPA requirement for generating those renewable information numbers. So that, in order to make this work, that sort of, there would have to be some sort of regulatory change or relief there. Yeah, I’m asking because I know since all of the ones that are moved within the US because of those EPA requirements. they have to be denatured but as you mentioned, there are some, when we export them that are undenatured so it seems that perhaps, other countries already realize that it’s not necessary to do that and as you said, there’s been no cases that have been shown of theft of this. Not that we’re aware of. Thank you, Mr. Stancil. So there’s so many things to ask about this report. It’s very interesting and well-written, Mr. Mike Hiller, as I see, thank you. I wanted to focus on some of the recommendations. We conclude that FRA inspectors didn’t report all of the defective crosstie conditions in the two years prior to the accident. We make a recommendation to FRA about training. So do you think that this recommendation about training, I mean, it addresses the shortcomings because assuming that the inspectors weren’t adequately trained to recognize crosstie problems, but what about other problems? I mean, such as complacency or just oversight and reporting. Will this type of training also address that or how do we hope to address that? So, I think, first I’ll start off by talking a little bit about the disparity in the application of FRA inspections between Region 5 and Region 6. And in the report, we show a data that suggests there was some difference between the use of the civil penalty to enforce or compel the railroad to make corrections. In trying to understand that a little bit further, we had some discussions with the regions about the use of this tool to compel a railroad to meet the minimum track safety standards. We learned that it’s a challenge, I think, for some of the regions because it’s about establishing a relationship with the railroad and you don’t always get good compliance with simply enforcing or using the civil penalty process over and over again. So that was one of the aspects and I think Joe and I, we had this discussion this morning and I think you can help me out a little bit. Yes, so I think the concern is we always wanna be careful that we’re not just comparing numbers because sometimes, things can be lost but in this situation, we see that they weren’t obtaining compliance and the non-compliance coupled with the fact that they weren’t using their enforcement options to compel compliance is what the concern was. So, it’s definitely a consistency to make sure that the inspectors are being trained, the compliance manual is being discussed and the use of enforcement options when necessary is more uniform across the regions. Thank you, Mr. Gordon, Mr. Hiller. I am over my time, apologies. Good job, thank you very much, Member Dinh-Zarr. Member Homendy. Thanks, and I wanna follow up on that. First, this was a great report. You all did a fantastic job, thank you. In January 2015, the FRA commenced a national program to inspect Class 1 railroad track where crude oil and ethanol were transported. Those inspections revealed 5,700 defects on UP’s system alone. No civil penalties were issued. From 2015 to 2017, FRA inspectors from Region 6 identified 32 defective crosstie conditions, some of these were also identified by UP. No civil penalties were issued. Then two months after the accident occurred, in May 2017, FRA again inspected the track and found 78 defective track conditions, 51 of which had defective crossties and five were locations with rail fasteners not maintaining track gauge. Again, no civil penalties issued. I’m not suggesting civil penalties is always the answer, it’s a tool especially when FRA, they issue a civil penalty and then it generally is negotiated at 50 cents on the dollar and I think looking at UP’s statistics for 2017, FRA assessed civil penalties for about 5.2 million and those were settled at about 3 1/2 million so it’s not always the tool but I do wonder what training or guidance the inspectors get for using civil penalty authority. Yeah, so before coming to the board, I was with the FRA as a track safety inspector on Region 2 and the training is very good. There’s track recurrency training and at the time, when you first on board, there’s even what they call a great inspector training and they talked to you about this exact thing. You do have to work with the railroad. It’s not always by using the big hammer but at the same time, they talk to you about progressions so once you’ve brought a location or an area of concern to the attention of the railroad, now, it’s time to monitor and see how they react to it and then you move to the next stage which may be civil penalty. They can do multiple counts, there are many different things that they can do depending on what the response from the railroad is so you so you kinda incrementally use your enforcement options. And that’s what we weren’t seeing here is that they were doing a good job of notifying the railroad and there is a requirement within the FRA for Track Safety Standards violations that there’s prior knowledge. So the prior knowledge is definitely there when for two years, they’re telling the railroad, you’re marginal here. You’re not gonna make it through another winter cycle then when they come back and find 78 defects postaccident, there would be an expectation of some civil penalty recommendations. Well, in looking at the inspection reports which are in the docket, there’s also a category on the report that where FRA would check written notification to FRA of remedial action taken by the railroad and in every instance on the inspection report, they checked optional, not required. You don’t have to tell us when you remediate the condition so that’s also an option that they have for determining whether the railroad actually took action after they identified a defect. So for Region 6, it’s comprised of Colorado, Iowa, Kansas, Missouri, Nebraska, Southeast Wyoming and some southern portion of Illinois. How many inspectors are in Region 6? I know that we looked at that number. Is it 16? I believe it was 18. 18? And that’s, of course, supplemented by the state agencies as well. Mm-hmm, and how many of those focus just on track? Do you know? I believe that that number is the track’s force. The full. Yes.
Okay, great. Okay, I’m gonna hold the rest of my questions until the next round. Thank you, Member Homendy. So according to the report, we’ve seen other, the report called out other accidents involving UP’s ineffective track inspection and inadequate management oversight including an Amtrak derailment in Arlington, Texas that was operated on UP track. Does anybody remember what year that might have been? It’s 1998, I believe. Okay. Then there was a UP train on UP track in Eunice, Louisiana and Mosier, Oregon in June of 2015 and now this one. Now, after the Mosier accident which involved a crude oil train and inadequately inspected and maintained track, the FRA entered a compliance agreement with the UP that came out, as mentioned, on December the 22nd of 2016. On page four, it says the FRA has conducted other inspections of UP track which coupled with UP’s track inspections, discovered as a result of the derailment described above, the Mosier derailment, raised safety concerns about UP’s track inspection program. This is the FRA’s compliance agreement with UP which UP signed and agreed to. It says the track inspection records audit also revealed deficiencies with UP’s records and it talked about record deficiencies. It also points out that UP was notified by the FRA 12 days before the FRA was gonna come in and do basically a blitz. So UP had the opportunity to get out there and correct many of these deficiencies. In total, UP’s track inspections documented 120 track defects in a 74-day period prior to the FRA audit and 181 defects during the 12 days before that. However, in spite of that advance notice, the FRA team found 664 track defects. So in other words, even with word that the FRA was coming on property and they had the opportunity to get out and correct those things, the FRA still found 664 track defects. And as mentioned, there was a program called the CORTEX program which revealed 5,700, 5,700 defects on UP’s system and FRA is considering enforcement action for more than 800 recommended violations. So when you see it as widespread as that in accidents, derailments across the country over a period of time, when you see that some of these inspections on the CORTEX, on this blitz were in Region 8 and some were in Region 6, I think. Region 5. It indicates this isn’t just a one-off problem. It’s a systemic problem. Somebody sent me last night a news release from April of this year saying that Union Pacific is the safest US railroad for the third consecutive year and they’re talking about employee safety, occupational safety which is very important but our job here at the NTSB is to look at transportation safety. Safety is holistic. You can’t just be safe in one area of your organization and not in another. And I think between this investigation where we found widespread deficiencies coupled with previous accidents, coupled with the FRA’s compliance agreement which UP agreed to, what does this indicate? Mr. Hiller. Chairman, it’s a clear demonstration that the state of good repair of the transportation infrastructure as a whole is in need of an infusion of capital, maintenance and attention. Please. And a management commitment to make sure that the inspectors have the proper resources to do their jobs but the FRA has a role in there as well because we found deficiencies that the FRA had gone and looked at this track but we still found deficiencies. I’ll get on to more of that later. We’ll begin with the second round. Vice Chairman Landsberg. Thank you, Mr. Chairman. What’s the life of a crosstie in Iowa, typically? Yeah, so that’s an interesting question. I’ve been doing a lot of reading on that. So, the crossties for a long time, 30 to 40 years service life. On thinking of a standard operation, not super heavy tonnage lines but 30 to 40 years service life. Over the years, that’s gone down. Most railroads use 25 to 30-year service life of a crosstie now. It just has to do with the quality of the wood, the quality of the treatment has probably gotten better but the wood that you start with is what you’ve got. So most of the railroads, their renewal cycles, they work out depending on the tonnage of the track but an industry standard on tonnage line like we’re looking at with the Estherville Subdivision would be somewhere between eight and 10 years to go in and replace one quarter of the crossties. They would go in and replace 25% of the crossties on that territory. Was that done in this case? No, sir. The last tie, the last production tie program that they ran was in 1987. So 31 years and that was, so this is what the industry refers to as deferred maintenance and so there were several tie renewal cycles that were deferred. That seems like a charitable description, sir. Is there a way to automate cross tie inspections, somewhat? It sounds like a fairly tedious job to go out and walk the tracks for miles and miles. Good for your physical fitness but may be a difficult way to catch things and I could see complacency with the track inspectors being a bit challenging. Right, the complacency is an issue. And so one of the ways that the visual inspection is supplemented is with the geometry program. So the railroads run automated cars that check the gauge surface and alignment and the curvature, the superelevation. It does all these at track speed. In addition, some of the cars are equipped with a gauge restraint measurement system which is like a split axle, so they’re actually shoving force on the rails to check the sturdiness, the robustness of the track structure. Can the industry meet the 350 car 117 requirements? And who’s really monitoring that? I noticed we have a recommendation R-15-16 is open and unacceptable. Where are we on that? So, in terms of monitoring the progress, at this point, it’s the Bureau of Transportation Statistics who publishes an annual report to Congress. They released– Does anybody read it? Well, they released, I read it. Okay, all right, I just. They released a report in September. We just recently received their report for, for 2017 and in addition the, AAR publishes statistics to the industry quarterly on the progress of the retrofit and status of the flammable liquid fleet. I’m sorry, what was the rest of your question? I’m just curious, is this a feasible goal? Can they meet the 350 car per month requirement or are we looking at getting to the end of the time here and people throwing up their hands and saying, well, this wasn’t realistic to begin with and who’s holding them account, who is holding the shippers, not the shippers but the people who leads the cars accountable? According to the Bureau of Transportation Statistics data, for 2018, they project 3,817 DOT-117s to be built for the year and 6,125 DOT-111s and 1,785 CP-1232s will be retrofitted to DOT-117 standards for a total 11,700 cars all together. For the ethanol industry, so to take a look at their fleet, it’s composed mostly of, originally, it was composed mostly of the legacy DOT-111 cars because it started the ethanol unit train shipments really ramped up in the mid-2000s before there were any CPC-1232 so their fleets were mostly the 111s. So to replace those cars would require either retrofitting the older 111s or buying the newer car. So if they were to buy newer cars to maintain that 350 car per month pace, that would essentially consume all of the cars, all the new cars that are being built based on the figures for this year. So that in itself is a concern plus you have the crude oil market and the other flammable liquids markets who are also in need of these tank cars. Thank you, sir. Thank you. Member Weener. In the last round of questions, we talked about failure of the rail itself. What part in the track failure does the crossties and the ballast have in terms of the track failure? So there have been a lot of studies on this as well and the thermal stresses are the number one leading, I guess promulgation to make rail defects, internal rail defects grow as well as residual stresses in the rail, track curvature, track modulus or the vertical stiffness of the track is further down the list of the leading, the things that contribute to the growth of internal rail defects. So, it was mentioned earlier that where on the track, on the rail itself is a important factor. Yes, that is an important factor because the head of the rail carries that load and the more area that can distribute that load, the higher the stress is. The less stress is on that rail. So as the rail profile wears down, the head area gets less and less and so while the loads may actually be the same, the stresses will continue going up in the rail until you start getting into areas where if you have an internal defect, it can begin to grow or if you just have, just meet the, basically the material, the material properties of that piece of rail. So the head of the rail is actually in compression. There’s a… Load rolls over it, is that correct? That is correct. So–
As it’s under compression, it has to be balanced out with the tensile to elsewhere too. Yes, because it’s essentially an I-beam. That is correct. So, 90-pound rail, is this was and it was considered to be light. After where it becomes something less than a 90-pound? Per three foot? You could conceptualize it that way, yes but that is correct. So what typically, in terms of the rail itself, what’s used elsewhere in UP’s system? So, the smallest rail that’s being manufactured today is 100-pound rail. The Class 1 railroads don’t use that. That’s used on commuter railroads, really much lower axial loads. Typically, anything that’s laid today is 141-pound rail, so a much larger rail section than the 90-pound. How much of the failure in rail itself is due to fatigue? Are you discussing this rail or all rail failures that we gonna get. In particular, this case. So, in this case, we were not able to determine or find a fracture surface that showed fatigue crack or any kind of growing defect that would lead to premature fracture. There were those features as I discussed on many pieces of rail particularly increasing as you get closer to the missing 12.6 feet of missing rail on the south rail. That’s where the most concentration of those pre-existing defects were located but none of them were large enough in and of themselves to cause a crack to grow to break the rail from the pieces that we’ve recovered. All right, I think I’ll yield the rest of my time. Okay, thank you very much. Member Dinh-Zarr. Thank you, Chairman. Mr. Stancil, so are we saying that at the current pace, the retrofitting and construction of replacement tank cars will not meet the deadline? Well, based on the figures for this year, they look very promising. So for 2018, just looking at the first half of the year, there was a remarkable number of cars added to the fleet. Currently, in ethanol service, there are 14,000 DOT-117s, about half of those are retrofits and about half are new cars. Compare that to 18,700 DOT-111s and about 3,000 CPC-1232s. Those are the ones that need to be replaced. So you say it is promising. So it is if we keep up at this pace. If the pace continues.
It’s is possible. Just the last several months does show some promising progress. Well, I hope that will actually happen. Thank you, Mr. Stancil. I know that this accident hasn’t gotten as much media attention because thank goodness, there were no injuries and there wasn’t a big fireball but I have a question about emergency response. Is there a difference in emergency response with denatured and undenatured ethanol? Especially for these, I mean, in this case, it’s a really small, probably a very small fire department. Yes, there was a volunteer fire department. So the DOT Emergency Response Guidebook refers first responders to the same information for both undenatured and denatured fuel ethanol which those guide pages in the response book are pretty generic and they’re intended to help the first responders get through the first phases or the first few minutes of an incident until experts arrive on scene but there are a few unique properties to undenatured ethanol that the first responders would need to know and for instance, the undenatured ethanol and we observe this ourselves, is that it burns with a colorless flame in daylights. It’s invisible in daylight. So, for instance, if the first responders were unaware of this fact, and they had to approach closely to the area to perform an expectation or an evacuation or take some sort of offensive action within the derailment, they could potentially be at risk. So, that’s an example of property of the material that they would need to be aware of. Thank you, Mr. Stancil. But essentially, their goal is to clear the area, is that correct? And it seems that they did what they were supposed to do.
Yes, this response went very well. So their first goal is, of course, beyond life safety and protection of life. We would expect the first responders to establish an evacuation zone of an appropriate distance, isolate the scene and gather as much information as they could about what was involved in the incident until the arrival of the railroad’s commercial firefighting service, hazardous material specialists and tank car specialists. We wouldn’t expect the first responders to manipulate or handle any of the rail equipment although sometimes, they might assist with that sort of activity once the experts arrived on scene or they have the appropriate guidance but in this case, they stood back. There wasn’t a need to approach closely because of the remoteness of the location. Thank you. I have two very short quick questions for Dr. Mueller about the rail. When we say rail weight, that is how much a three-foot section weighs, is that correct? That’s correct. Okay, so the rail that failed in this accident, I know there’s been talk about the different weights. A 90-pound versus 112 versus 115 and UP had installed heavier weight rail in other locations but do you, in your analysis, is the presence of this lighter weight 90-pound rail, was it a factor in this accident? So first, I’d be remiss to not give credit to Xiaohu Liu who actually did the analysis but what the analysis did, er, sorry, what’s the finite element modeling did show was that there was nothing where the stresses were gonna exceed the properties of 90-pound even with the less stiff track support in and of itself. That does not preclude that the track condition may be worse, that there might be pre-existing defects in a piece of rail, things of that nature. Thank you, Dr. Mueller, I just wanted to clarify that. In this case, it was unlikely that the weight, a heavier weight would have done much more than the 90 weight since it didn’t show much, I mean in your analysis. I know it’s a little bit more complicated than that but– If your question is that the performance may have been enhanced had there been a higher weight rail, that is possible but that was part of the thing is we wanted to see if there was something inherently wrong with the 90-pound rail that it was not robust enough for current heavy freight loads. And you said it was. It was. It was robust enough. It was.
Thank you. That was my question. Thank you very much. Sorry to go over my time. And Member Homendy. FRA collects safety data from the railroads that they publish on their website and then in that safety data, they talk about train accidents, grade crossings, employee on duty cases, trespasser fatalities and they break out the causes of train accidents, not including grade crossings. What’s the number one cause of rail accidents? Train accidents, not including grade crossings. This bounces back and forth between track and human factors. It’s human factors, right? So I looked at the last 10 years of data for the seven Class 1 railroads. All number one human factors except for UP. Every year, it’s track defects. So that would suggest that there is a problem with their track maintenance program that’s goes back quite a bit of time. And I just wonder, since you were with FRA, do they analyze that data that comes in from the railroads? That should be used to drive their inspection programs, to have them focus their limited resources on which areas they need to send their inspectors. Okay. So for this subdivision had 79 miles of track and from 2015 to 2017, UP identified 102 track defects on that stretch, noting crossties that were either marginal or in poor condition. In the regulations, what do the regulations require for remediating a defect? And I’m gonna ask specifically, I think there’s two ways. It’s when they take it out of service but there’s also when they allow transportation to continue on that track. Depending on the type of defect, there are class specific defects and non-class specific defects and speed defined defects but that’s a very small number. So if it’s a class specific defect that the inspector finds, then he has to immediately reduce the speed of the maximum authorized speed on that line to meet that class of track. Then in addition to that, the remediation, if it’s a non-class specific defect, then they have up to 30 days to make that correction without reducing the speed of the track and when you get into crosstie defects, you have some of both and a defective crosstie, it just depends on what condition is found. And this was Class 3 track which is rated for 40 to 60 miles per hour but they speed restricted it to 30, right? That’s correct. So obviously, they knew there was a problem in this area if you’re gonna speed restricted. Yes, I believe. Mike, did you speak to UP about the reason for the 30 mph? The 30 miles per hour imposed by the UP was, in their words, it was just really related to a timetable. They weren’t restricting the speed due to deferred maintenance. Oftentimes, when the ties are found defective like you saw in those so many accident reports, you saw that there was a remediation maybe a week later and what does that really mean? Did they go in and replace every single defective tie that they found? Likely not. In a 39 section or a 39-foot section of track, if you’ve got defective ties, they’re gonna go and restore to the minimum standards so maybe they have to replace five or six ties which would account for the overall condition of the Estherville Subdivision. Poor tie condition. Okay, and let me ask you about that with the inspections. In some of these cases and we condensed some of the inspections and remediation, at least, well, in several areas and just for an example. In July 2016, UP identified defective crossties at milepost 73.12 and stated they repaired them a few weeks later but in August, they identified the same defect. So what are their maintenance records like? Yeah, that could very well be just because of the reporting, like Mike was mentioning, if an inspector comes in and writes a location for crosstie distribution, then he will also mark the location. They may come out and slide one tie in that location and call that remediated because now, it meets the minimum number of effective crossties in a 39-foot segment. So then on the next inspection, coming right off of that new installation, there could be another distribution defect and there again, he’ll send the maintenance crew out to that same location. Okay, great, I have additional questions but I’ll wait. Great and just for planning, after I’m completed, we’ll plan up break right after this round. So can you specifically enumerate the rail shortcomings that this investigation found postaccident in the Estherville Subdivision? Track structure defects? Yes, please. I can, I know on the walking inspection, we found, I believe it was seven locations and that was just right close to the area of the derailment. All right, that’s fine. If you have a more robust list, then we can do it after the break but if not, if that’s really all you have, then that’s fine but… Tell you what, I’m going to defer that question until after the break and also, I’m going to wanna know and after the break what specifically, okay, so we identified these things postaccident but then the FRA, did they overlook or somehow not observe these issues? So that’s what the question will be. I’m just hanging that up for after the break. Member Dinh-Zarr asked about if the 90-pound rail itself might have been a factor here and I think Dr. Mueller said no, it was 90-pound rail in itself is okay and I think Member Weener asked about the age of the rail. So if we had to say, was it more of a rail defect or more of a track support issue with the defective crossties and all of that, which way we’re leaning? So one of the things that’s difficult in this investigation is that we’re missing 12.6 feet of rail but one of the things that is reading between the lines for the finite element modeling study is that inherently, this piece of 90-pound rail should have been okay in and of itself even considering some deteriorated track support but that does indicate that there had to be something wrong with that piece of rail, the pieces of south rail that likely the ones missing and the other issue is that, again, is that we saw the amount of defects, pre-existing defects increased as one got closer to that section of missing rail and that piece of rail, that missing rail is located is at the transition from one type of fitness, the ballast to another type on the bridge and if you were looking at the recorder video, you even saw as the locomotive passed over the bridge that it did hit a jolt. So all these just factored into that, the likely point of derailment in there. One thing we do know is that there was a substantial number of rail ties that had were deteriorated and we have a picture of that, slide 26. We don’t have to look at it right now but slide 26 was a good example of that. You saw cracking in the tie, you saw movement of the tie plate and you could tell that because the tie plates were sort of cattywampus to each other. The ends of them weren’t all parallel, there was some burning where the tie plate had moved. So we have direct evidence of that. Dr. Mueller, well, I have you on the hook. I think what you may have said in your presentation was that we know that the derailment must have occurred, I’m sorry, that the rail break must have occurred prior to or at the 20th car. Did you say that? That’s correct. In reality, the physical evidence might indicate that it occurred as early as the fourth tank car. Is that correct because we saw those, I’m gonna call it batter. Better is more the function of the rail itself but we saw the banging, some evidence. Yes, those impact marks on those cars are consistent with the rail having broken as those cars were passing by. Thank you. Okay, I tell you what we’ll do. We’ll take a break. Let’s come back, according to that clock, let’s come back at 11:30 and we are in recess. We’ll begin in about two minutes. Okay, we’re back in session. We’ll begin with the next round of questions beginning with Vice Chairman Landsberg. Mr. Hiller, could you describe the relationship between the FRA and the railroads in terms of this compliance agreement and kind of how that whole situation works to make sure, ensure that, in fact, the standards are being met? Well, first, it’s important to understand that within the FRA, there about 400 inspectors across the United States. The compliance agreement was a tool. A tool that through partnership between the regulator and the railroad that was developed to improve conditions to ensure safe travel of high hazard flammable material trains. Mosier, Oregon was the result of a wide gauge derailment and they, meaning the FRA, had a problem with the way the rails were fastened to the track and as our chairman pointed out, the context of the compliance agreement, there were these CORTEX inspections and they really blitzed the Union Pacific Railroad and they found a number of problems with the way things were being maintained. So how best to address them would be the question. They came up with this agreement and although the railroad didn’t agree with all of the findings from the FRA, they did agree to the compliance agreement which meant they agreed to meet certain expectations laid out and there were milestones and there were timelines associated with meeting those milestones and the overarching goal was to improve the conditions of the infrastructure to ensure safe movement of hazardous materials. Joe, can you add anything to that? Yes, typically, the compliance agreement would be, and it’s referred to in the Code of Federal Regulations as a compliance order. Typically, the compliance order would be if there’s a large system wide issue, if there’s an issue identified on a certain territory like a subdivision like in this case then the FRA also has the ability to do a special notice for repair. Special notice for repair, the FRA can decide what restrictions they’re gonna place on the railroad and that could be run at 10 miles per hour and then there’s a list of defective conditions that need to be repaired and someone from like the engineering, higher level manager and engineering has to sign off that all those repairs have been made. So the compliance agreement is like a step above special notice to repair and dealing with more of a system-wide issue. In your opinion, do you feel like the spirit of the compliance agreement was met here? It was kind of–
I realize I’m putting you on the spot here but. Yeah, it was kind of early to assess it because it went into effect December of ’16 and then the accident was in March of ’17 so we really didn’t have time to assess how effective the compliance agreement was. The data that we have now suggests there’s been an overwhelming response to the compliance agreement. The chairman mentioned and so does the report, there were more than 8,600 crossties that were changed in just the Estherville Subdivision alone. The target was many of the ethanol, those short ethanol routes that are in the Iowa in that Region 6 area were also gonna be part of campaign by the Union Pacific. And the report reflects that there was what we would consider to be very good response to the compliance agreement. Well, that’s encouraging. I guess I would just say we need to be very careful not to get too complacent. Thank you. No further questions, Mr. Chairman. Thank you very much. Member Weener. Thank you. We’ve talked a bit about the characteristics of the roadbed and in fact, that has on the life of the rail. The accident or the initial point of derailment was route the western approach in a transition area to the bridge and we haven’t talked much about the bridge. Is the stiffness of the bridge an issue in terms of the life of the rail? Yes, that’s a very demanding location for maintenance. Railroads have identified that as being an area that requires additional attention. In addition to just doing more surfacing and paying closer attention to the surface off of the approaches of the bridge, railroads are now installing 10-foot crosstie, not now, they’ve been doing it for quite a number of years installing 10-foot crossties to help to, to help to defer some of that load as they’re approaching the more rigid bridge structure. So it’s intended to better distribute to load. Are we able to really inspect any of the components of the bridge? No, sir, because of the trailing tonnage, the cars that came in on the derailment area and powed up there as you saw in the animation and the fact that the bridge caught fire, the bridge structure caught fire. There was no component. We had to go off of the last bridge inspections in the interview with the bridge inspector. So what kind of inspection was the most recent than the inspection? Was it one of the automated inspections where you run the car down or was it of walk over? The most recent, I believe, was the day before the accident and it was a high rail inspection which would be high rail walking. If they identify anything from the high rail vehicle, then they stop and get out. But what is, actually the issue here is the vertical loads and the movement with vertical loads, is that correct? That’s correct. And does the high rail actually represent anything near those loads that the actual locomotives and tank cars in this case would out on the rail? No, sir, in this case, the visual inspection, the inspector would be looking for crosstie hanging, looking for the location where the crossties actually swinging below the rail and then at that point, he would get out and take a measurement with the tape measure to see how far down the rail was going under load. So, there’s really a good means to automate the inspection of the ability to take a vertical load. No, what we hope would they would be doing is supplementing the visual inspection with the last geometry car inspection, so looking at the data from the last geometry car inspection and focusing their attention when they’re doing their inspections on areas of concern but the inspector was interviewed and he said that he had no concerns on that bridge end on the last inspection. When you say geometry car, is that the same thing as the high rail? No, sir, no. The geometry car is either a self-propelled rail car that is equipped with all the testing equipment or some of the cars are pulled, coupled to locomotives and pulled across the territory. So the geometry car does or does not put a representative load on the rail? The geometry car does. It does? Yes. And we have data from their recent inspections? Yes, I would have to look at the date of that most recent geometry inspection but there were no defects noted in that area. But that would have indicated if there were a problem with the transition area? It would have identified a defective condition. If there would have been a condition a geometry condition just starting to manifest until it gets to a defective condition, sometimes, they’ll test one track level above. Yes, the most recent was in May 24th. Oh, that was the rail test, yeah, okay. Yeah, the August of 2016 was the geometry inspection prior to. All right, and one last question. I ran across the terms checking, shelling and flaking. And I thought maybe I was getting a description to a hockey game. What are those terms? So these are all terms that are used in the rail industry. Most of them are based on the orientation of basically defects that are growing crack, cracks are growing as a result of rolling contact fatigue. So flaking is actually like one of the least serious ones but basically, it’s just a, it’s a horizontal separation on the running surface of the rail. Head checking are transverse surface cracks. They’re on the corner where the wheels being applied and those are transverse going downward to the rail and then shelling is another type of subsurface progressive cracking. That’s more longitudinal in orientation. Again, along the running head of the rail surface but all these are products of rolling contact fatigue that starts small initially and then develop over time and grow. All right, thank you very much. No more questions. Thank you, Member Weener. Member Dinh-Zarr. Thank you, Chairman. I have one last question and it’s for you, Mr. Hiller. We make some conclusions about Union Pacific Railroad so would you tell me about UP’s safety management system and just describe that and tell us more about it because it seems with an SMS, a lot of these problems would have been prevented. So SMS is often discussed here at the safety board and I can tell you that the Union Pacific had elements of the SMS program. They had policy, they had risk management, they had the assurance and they had the promotion. So let’s talk about the Estherville Subdivision and how the application of this type of a practice was designed or used to find some of the conditions in the track. The inspectors were going out and they were doing their inspections and they were reporting the conditions of the tie, marginal, defective and in some cases, and in many cases, more than a hundred, as Member Homendy pointed out, there were remediation efforts and it’s clear based on our observations postaccident that the remediation efforts restore the track back to its minimum condition that it needed to sustain traffic. The regulator went in providing this assurance or providing this oversight and again, he would comment that there was a condition and they were concerned about it and were moving ethanol trains. So there were some elements there of an SMS system. They had a policy, they had their minimum safety standards outlined in the regulations and the Union Pacific also had their own maintenance and track standards manual. Now, we get into the recommendations where we were asking the Union Pacific to take a look at these things. What are we really looking for? We’re looking to see that things are not just restored back to the minimum. We know that doesn’t work. If you’re finding 10 or 12 crossties in a 39-foot section of track that are defective, it’s not a good practice to go in and replace two or three just to restore the track. Perhaps, you have to do more than just the minimum and I think that’s the message here. So when I speak to the Union Pacific’s SMS program, I think they have elements of it. Is it formalized much like the board has discussed in past investigations? Likely not. The Federal Railroad Administration is on the hook to produce Part 271. This is a risk management approach and it’s just that. You make a determination of how much risk you have and whether or not it’s acceptable. But this alone isn’t without risk. Thank you, Mr. Hiller that was very helpful and I am sure that, I’m confident that this accident was a wake up call for UP and whatever they end up calling it, I’m glad that there are elements of it and I hope that there will be more elements, so thank you very much. Thank you, Chairman. Thank you, Member Dinh-Zarr. Member Homendy. Dr. Mueller, I don’t have a question for you but it’s my understanding that you recently received the Silver Medal Award from ASM International for your work on failure analysis so I just wanted to say congratulations on that. I have two questions, probably, for Mr. Stancil, I think. First has to do with firefighter training and the fire chief had specialized training. I know that Title 29 Code of Federal Regulations, Part 1910.120 sets forth different levels of training for firefighters dealing with hazmat response. General awareness, operations level, technician incident commander. And most firefighters, unfortunately, are provided with general awareness training. This fire chief had specialized training above and beyond what’s required in the regulation so I was wondering if you could talk about that training and also talk about whether that training is available for other responders. So, this training that you’re referring to referenced in the OSHA HAZWOPER regulations is discussed in the, I think it’s NFPA 472, standard 472. So there are a number of items that the firefighters but the train of that level would need to know certain things such as predicting the outcomes of a particular incident, knowing how to obtain the information about the hazard and applying it to a specific situation, getting into remediation or taking offensive actions. There’s other levels of training. I’m not completely aware of that the training that this firefighter received other than the specialized courses that he received, an ethanol emergency response which were sponsored by the UP. I think he and a number of his staff had attended two courses that were provided post that initial training that he received but he was very familiar with what was traveling through his jurisdiction. In fact, he said that we see ethanol trains here every day so it wasn’t any surprise to him that that’s what he was dealing with although initially, he was unaware of there being any difference between denatured and undenatured ethanol. Mm-hmm, thank you. And I want my last set of questions is on buffer cars. FRA regulations require that placarded cars be placed no closer than the sixth car from the engine but only when train lengths permits. My question is what does when only when train length permits mean and do current FRA rules provide an adequate buffer between hazmat cars and the locomotive, if not, what recommendations have we issued on this? Okay, so you’re referring to the PHMSA rule in Part 174. It’s a regulation 174.85 regarding the buffer car requirements. So as a result of an accident that occurred back in 2006 in New Brighton, Pennsylvania, the NTSB approached PHMSA and requested a regulatory interpretation of this regulation because at that time, railroads were interpreting that requirement that mean that they did not need to add additional freight cars to a train if none were already contained in the concept. And so PHMSA’s interpretation in response to the NTSB in 2007 essentially said that the five buffer car requirement applies so long as there are sufficient non-hazardous material rail cars within a standing train consist to fulfill the requirement and that the regulations do not require railroads to change business or operating decisions concerning the number types of cars placed in a train. So they do not require the railroad to go out and find additional buffer cars that a shipper might not have. So, in this situate, you had a loop track, right? So you had two locomotives of buffer car, 99 tank cars, another buffer car and another rail locomotive. So it’s the issue that you can’t put five on because you’d have to take five tank cars off? Yes, so we’ve heard this argument. The FRA held a Rail Safety Advisory Committee meeting on hazardous materials and this issue came up and that was one, I’m not sure if that was the case with this specific shipper but shippers have expressed that concern that loop tracks are spaced to handle a certain number of rail cars and if you have to add five buffer cars, that’s five less tank cars that they can add to a train. So that was one reason they were not in favor of requiring five buffer cars but our recommendation out of the Casselton, North Dakota accident, we issued that report earlier this year and we recommended that PHMSA withdraw that interpretation until they determine what the, and working with FRA until they determine what the appropriate separation distance should be between an occupied locomotive and the first hazardous materials car. Did they do that? That recommendation is currently classified open with an acceptable response. They have not withdrawn the interpretation. Okay, great, thank you. Thank you very much, Member Homendy. Mr. Gordon, do you have those figures I was asking for? Yeah.
Go ahead. So what deficiencies did we find postaccident in the Estherville Subdivision? Okay, so just to let you know a little bit about what we did and what we didn’t do. We only inspected, accompany the FRA for about one quarter of a mile. They inspected about a half a mile of track. Jim Southworth and his mechanical team called me and said we’ve got witness marks on the wheels so the track inspection then became focused on rail recovery but yes, we did, in the a little under half a mile that we looked at the day after the accident, seven defects were identified. They were four crosstie distribution defects, one insufficient number of crossties in a rail segment, one rail fastener defect and one concentrated load between the rail base and the top plate was what was identified. When you talk about crosstie distribution, what was the term used for the first one, cross errors? Yeah, crosstie distribution and then an insufficient number of crossties in a rail segment. So those are the class specific and non-class specific defects that I spoke about earlier. A crosstie distribution defect is a non-class specific defect where an insufficient number of crossties in a 39-foot segment requires a speed restriction. Thank you very much. So, the UP is a requirement of the compliance agreement that was entered into about 2 1/2 months prior to this Graettinger derailment. It required the UP track to be inspected in certain classes twice a week. Now, was that done? Did we find evidence of that being done twice a week in this subdivision? Yes, sir. Before the compliance agreement, it was once weekly was the inspection frequency and then following the compliance agreement, I think there was only one week within the time that we looked at that they did not make the twice weekly inspection. But they were required to do it twice a week by the by the compliance agreement. There was one week, the week of April, February the 19th that they did not do it twice a week as required by the compliance agreement. Correct. What are the implications of that? At a minimum, it would be a defect taken to the railroad and it could rise to the level of a recommendation for civil penalty depending on the inspector. Because the terms of the compliance agreement, wherever it is, were that UP had to comply with those or the FRA would get a little bit more heavy-handed. Isn’t that correct? That’s correct. Okay, thank you. A little while ago, the vice chairman asked a question, rhetorical question, wouldn’t it be a good idea to have different grades of class according to what’s being carried, that might have been your question and in fact, there are different classifications, for example, there’s a key routes and there’s routes that are HHFT routes, I think. Can you discuss that to to enumerate how there are different routes according to the product being carried? Yes, so I believe Paul can speak more to those classifications but I did want to clarify something. I’m glad you brought this up. When that question was asked, I was thinking more of crossties and gauge and alignment and those things. In the 213 standards dealing with internal testing of rail, there is a requirement that on the high hazardous flammable train routes, the defect ratio has to be kept lower or they have to do more testing. So, just to clarify that point and then I believe Paul’s ready for the– Just a thumbnail. Yeah, so key routes and the concept of key trains is an industry classification. So the AAR has developed this requirement and it has to do with frequency of track inspection, I believe, and there are some other operational restrictions and requirements but, and it has to do with tonnage over a certain territory and over a certain period of time. So, key train, the definitions have changed a bit over the years but essentially, I don’t have it right in front of me but it’s dependent on the number of hazmat cars in the consist. Yes, and for certain routes, high hazard flammable train unit trains, there’s a minimum of 27 safety and security factors based on the route. That’s correct. And so there are differences but the reason this was not in that classification is because of why? They were hauling ethanol but it was because there was only one train every other day. It was tonnage over the line, yes. Thank you. I have a few more questions but I wanna make sure that my colleagues don’t have any. I do wanna comment. I’m curious I thought the animation was remarkable. That was the first time I had seen it. I wanted to see it again because I thought it was, there’s a lot to watch. The forward-facing video plus the actual animation. Really a nice job with that, Barbara, to your team. And well, you came out of research and engineering, do you? Yeah, but based on the data from the actual field investigation but how did the drone, NTSB operates a drone. So how did the drone assist us in this investigation? It’s not just a toy. Yeah, so the drone work, Bill English and his team assisted us with that. They operate the drone and so, do some of the railroads as well. The drone helps us immensely in that it documents the resting position of the wreckage, the tank cars. Previously, we had to walk through the wreckage, do it by still photography, diagram and out and it was a laborious and time-consuming process and imagine doing that on a rail line that’s a major route and the rail carrier is interested in getting it open as soon as possible. Well, that pressure wasn’t there specifically for this case but in other cases, it definitely is. The drone work helps us speed that whole process up. That gives us the context for the damage assessments of the tank cars so we can understand better car to car interactions and help us in our analysis on what caused breaches in certain tank cars and that sort of thing. It’s essentially become an essential tool for us now. It is indeed a tool and like I said, it’s not just a toy. We found it and useful in many applications, many investigations. So I guess, Mr. Stancil, I’m going to ask this to you. Back in 2013, of course, you worked with the Transportation Safety Board of Canada with the tragedy at Lac-Megantic then several of us been a cold New Year’s Eve in Casselton, North Dakota. Both crude oil train, crude by rail accidents and we were very concerned about crude by rail, even the phase out for the DOT-111s put more emphasis on the crude by rail instead of ethanol because crude by rail, the DOT-111s had to be phased out by this year, whereas we have five more years for ethanol. As it turns out, apparently, the ethanol, denatured ethanol poses a greater risk to safety and more volatile than the most volatile raids of crude. Can you comment on that? Yes, that’s according to FRA research on this issue. They have determined that ethanol is in fact 1 1/2 times more, the properties are more hazardous and we get into the weeds on that as to why they think that but they looked at various properties of ethanol versus crude oil in terms of vapor pressure, ignitability, the ignition energy, there’s sort of various parameters that are examined here. They also looked at accident outcomes. Some of the more violent tank car separations and fireball eruptions have occurred with the ethanol derailments and I’m thinking of one in particular at Arcadia, Ohio. Well, thank you very much. Any questions, further questions from our colleagues? With that in mind, Mr. Jones if you’d please read the proposed findings. Thank you, Mr. Chairman. As a result of this investigation staff proposes 16 findings. Number one. Union Pacific Railroad was not maintaining the track structure on the Union Pacific Railroad Estherville Subdivision in accordance with Federal Railroad Administration minimum Track Safety Standards or its own internal track maintenance standards. Two, Union Pacific Railroad supervisors and managers were not ensuring defective crosstie conditions were being identified, reported and remediated in accordance with Union Pacific Railroad track maintenance standards and Federal Railroad Administration Track Safety Standards. Three, Federal Railroad Administration inspectors did not report all defective crosstie conditions on the Union Pacific Railroad Estherville Subdivision in the two years prior to the derailment. Four, Federal Railroad Administration inspectors were not using all available enforcement options such as a recommendation for civil penalties, to require Union Pacific Railroad to comply with the Federal Railroad Administration minimum Track Safety Standards on the Union Pacific Railroad Estherville Subdivision. Five, based on the observation of the fresh horizontal impact damage observed on the wheel tread of the fourth through the 20th non-derailed cars, examination of the rail recovered from the accident, the conditions of the crosstie structure on the Union Pacific Railroad Estherville Subdivision, the train likely derailed from a broken south rail that occurred prior to or at the 20th car of the Union Pacific train UEGKOT-09 as it was traveling over the west approach of the Jack Creek Bridge resulting from Union Pacific Railroad’s inadequate track maintenance and inspection program and Federal Railroad Administration’s inadequate oversight of the application of Federal Track Safety standards. Six, a deteriorated track structure, such as worn rail and degraded track support condition will cause more track movement and higher stresses in the rail. Seven, the finite element study demonstrates that worn 90-pound rail on a degraded track support will result in higher transverse stress in the rail head region exposing the rail to increased risks of failure due to a vertical split-head failure mode. Eight, based on federal research and observed accident performance of tank car head protection systems in this accident, it is likely that had the legacy US Department of Transportation Specification 111 tank cars involved in this accident been replaced with US Department of Transportation Specification 117 tank cars equipped with head shields, breaches and punctures which resulted in the loss of hazardous material from six of the tank car heads could have been mitigated or prevented. Nine, tank car shell puncture resistance improvements required for new or retrofitted US Department of Transportation Specification 117 tank cars transporting flammable liquids that are scheduled to replace the existing fleet of US Department of Transportation Specification 111 ethanol tank cars by May 1st, 2023, could have mitigated and might even have prevented some of the tank shell breaches from six of the tank cars involved in this accident. 10, if the tank cars involved in this accident had been retrofitted or replaced with US Department of Transportation Specification 117 compliant tank cars, the breaching damage to the top fittings of the 21st and 25th tank cars could have been avoided. 11, ethanol would not have released from the 22nd tank car had it been equipped with a bottom outlet valve operating mechanism that was designed to prevent actuation during an accident scenario. 12, since the Pipeline and Hazardous Materials Safety Administration has not established a clear set of intermediate metrics for evaluating tank car conversion and replacements, achievement of the deadlines may be overly reliant on future market and economic conditions. 13, given the lack thermal damage to tank cars or energetic postaccident fireball eruptions and less environmental impact observed in this accident compared with similar railroad accidents involving denatured fuel ethanol, it would appear a safety benefit could be derived from transporting ethanol without the use of volatile denaturant chemicals. 14, more research should be conducted to determine whether operational changes to shipping ethanol in its undenatured form would improve safety. 15, the erroneous shipping documentation identifying the hazardous material as denatured alcohol, instead of undenatured ethanol, did not have any adverse impact on the emergency response to this accident. And finally, number 16, none of the following were factors in this accident. One, the mechanical condition of the train to include the train’s braking system. Two, the performance of the train crew. Three, cell phone use by the train crew. Four, alcohol or other drugs by the train crew and five, the emergency response. Thank you, Mr. Jones. Are there any amendments for many of my colleagues? Okay, I have a motion I’d like to offer. This is the same, it was whatever was emailed out yesterday and it’s been distributed. My motion is for finding 13, amended to read as follows. Given the minimal thermal damage to tank cars, comma, lack of energetic postaccident fireball eruptions and less environmental impact observed in this accident compared with similar railroad accidents involving denatured fuel alcohol, it would appear a safety benefit, it would appear a safety, it would appear a safety benefit could be derived from transporting ethanol without use of volatile denatured chemicals and that is my motion. Is there a second? Second. It’s been moved, it’s been seconded by one of my colleagues, Vice Chairman Landsberg. And as far as discussion is concerned, when I read the word as it’s currently written, given the lack of thermal damage, there was in fact one tank car that did receive small thermal share, small thermal shells tear which was line number 33 as outlined on page 18 and just for the sake of accuracy, when I looked up the word lack, it indicates the absence of. So it might be a little nitpicking but when I read the finding again after reading the report, I said wait a minute, I thought there was one that ruptured and there was. So that’s my motion. Any questions, discussion? Mr. Hall, what would be the staff’s position on this? Staff concurs with your change. Okay, thank you. Any further discussion? It’s been moved and seconded to adopt to revise finding number 13 as read. All in favor, please signal with hand and say aye. Aye. Opposed, there are none. The motion carries unanimously. Now, is there a motion to adopt the findings as revised? It’s been moved by Member Weener, seconded by Member Homendy. Any discussion? Seeing none, all in favor of adopting the findings as just revised, please signal with a hand and say aye. Opposed, there are none. The motion carries unanimously. The findings are adopted. Mr. Jones, if you’d please read the probable cause. As a result of this investigation, staff proposes to following probable cause. The National Transportation Safety Board determines that the probable cause of the derailment was a broken rail that occurred as the train was traveling over the west approach of the Jack Creek Bridge resulting from Union Pacific Railroad’s inadequate track maintenance and inspection program and the Federal Railroad Administration’s inadequate oversight of the application of Federal Track Safety Standards. Contributing to the consequences of this accident was the continued use of US Department of Transportation Specification 111 tank cars. Thank you, is there motion to adopt the probable cause? It’s been moved by Member Landsberg, Vice Chairman Landsberg, seconded by Member Homendy. Any discussion? There’s none, all in favor of adopting the probable cause as read, please signal with the hand and say aye. Opposed, there none. The probable cause is adopted unanimously. Mr. Jones if you’d please read the proposed recommendations. As a result of this investigation, staff proposes three new recommendations. One recommendation to the Federal Railroad Administration. Recommendation one. Provide additional training to all your track inspectors on regulatory Track Safety Standards compliance and provide guidance of available enforcement options to obtain compliance with minimum Track Safety Standards when defective conditions are not being properly remediated by railroads on all routes that carry high hazardous flammable materials. One recommendation to the Pipeline and Hazardous Materials Safety Administration. Recommendation two. Sponsor research to study and publish the difference in characteristics between denatured and undenatured ethanol and the benefits that could be achieved by transporting fuel ethanol without the use of volatile organic chemical denaturants. One recommendation to the Union Pacific Railroad. Recommendation three. Reexamine your track maintenance and inspection program standards on all routes that carry high hazardous flammable materials and ensure those track inspection standards are complied with by both track inspectors and track supervisors. Staff proposes reiterating one previously issued recommendation to the Pipeline and Hazardous Materials Safety Administration recommendation R-15-16, that is require an aggressive, intermediate progressive milestone schedule such as a 20% yearly completion metric over a five-year implementation period for the replacement or retrofitting of legacy DOT-111 and CPC-1232 tank cars to appropriate tank car performance standards, that includes equipping these tank cars with jackets, thermal protection and appropriately sized pressure relief devices. Thank you, Mr. Jones. Is there motion to adopt the recommendations as proposed? It’s been moved by Member Dinh-Zarr and seconded by? I second. Vice Chairman. Any discussion? I see none. All in favor of adopting the recommendations as proposed, please signal with a hand and say aye. Opposed, there none. The recommendations are adopted unanimously. Is there a motion to adopt the report as revised? We just revised it with the revision to finding number 13. Is there a motion to adopt the report as revised? Member Weener has moved, Member Homendy has seconded. Any discussion? There’s none. All in favor of adopting the report as revised, please signal with the hand and say aye. Opposed, the report has been adopted unanimously. Do any members wish the right to reserve, any members wish to reserve the right to file a concurring or dissenting statement? Member Homendy. Anyone else? Seeing no others. Is there any further discussion regarding this report? Well, in closing, I wanna thank the staff for your hard work. I had the easy job. I came in on Saturday and left on Sunday. You guys were out there in the blizzard. Peter, you were with me too. So thank you. I wanna thank my colleagues too for the preparation and going into this meeting. Each of us has met with staff and ahead of time individually to offer comments and so thank you for your good prep and your debate and discussion this morning. Mike Hiller, I wanna thank you as the investigator in charge but I also recognize that nothing here gets done by just one person. It truly does take a team, an investigative team that includes research and engineering. It all takes a team but I’ve also realized that nothing gets done without the support folks, the people that pay the bills, that can fix the computers, that hire the people. It is truly a team effort so on behalf of the entire board, a sincere thank you, not only to the investigative staff but to the support and program staff as well. The recommendations just issued, if acted upon, will result in better compliance training for FRA track inspectors and clear guidance regarding available enforcement options. They will result in UP reexamining its track maintenance and inspection program standards. Today’s recommendations will result in research by PHMSA into whether alcohol, ethanol, rather, should be transported in an undenatured state. Furthermore, a recommendation first issued in 2015 and reiterated today, will result in progress milestone schedules for the phasing out of DOT-111 tank cars for ethanol service by 2023, if acted upon. Otherwise, we risk a so-called sudden realization that isn’t sudden at all. We could have seen this train coming down the track so we wanna mitigate that problem and make sure their reporting requirements in place transparency to make sure that the tank car owners and shippers will have time to meet the congressionally mandated phase-out. Anyway, with that, thank you. We stand adjourned.