Greenberg Traurig, LLP — Hazardous Materials Safety Interpretation

Page 1U.S. Department of Transportation 1200 New Jersey Avenue, SE Washington, D.C. 20590 ipeline and Hazardol Administration aterials Safel OCT 30 2007 C. Allen Foster Ref. No.: 07-0169 Suite 500 Greenberg Traurig, LLP 800 Connecticut Avenue, NW Washington, DC 20006 Dear Mr. Foster: This responds to your July 31, 2007 letter concerning regulatory compliance issues associated with the attachment of mounting pads for internal baffle support clips on MC 331 cargo tank motor vehicles manufactured by Trinity Industries, Inc. (Trinity). Specifically, you ask us to reconsider our May 2, 2006 interpretation (Ref. No. 06-0046) on this issue. That interpretation states that, in accordance with § 178.337-3(g)(3) of the Hazardous Materials Regulations (HMR; 49 CFR Parts 171-180), the welding of any appurtenance to the inside or outside of a cargo tank the lading retention integrity of the cargo tank if any force less than that prescribed in § 178.337- wall must be made by attachment of a mounting pad so that there will be no adverse effect upon 3(b)(1) is applied from any direction. We continue to believe that our interpretation is consistent with the language and intent of the We have reviewed our May 2, 2006 interpretation and the information provided in your letter. regulatory requirement set forth in § 177.337-3(g)(3). However, you assert this requirement is tanks, constructed with the internal baffle support clips welded directly to the cargo tank shell not needed to ensure the integrity of a cargo tank as Trinity has manufactured thousands of cargo without pads, without cracking of the shell material or cracking of welds resulting from attachment of the baffle support clips. Based on Trinity's experience, we plan to reassess the requirement in § 177.337-3(g)(3) as part of an upcoming cargo tank rulemaking. Given the circumstances outlined in your letter, we have determined that Trinity cargo tanks constructed with internal baffle support clips welded directly to the inside surface of the cargo tank shell should be permitted to continue in operation, subject to provisions of a special permit he requirements in § 177.337-3(g). The special permit - termed a "manufacture, mark, anc ale" special permit - would establish conditions under which the cargo tank motor vehicle. 178.337-3(9) 070169#

Page 2at (202) 366-4511. contact Delmer Billings, Director, Office of Hazardous Materials Special Permits and Approvals, If you have further questions, please do not hesitate to contact this office. Sincerely, dI. Mazzullo Edward T. Mazzullo Director, Office of Hazardous Materials Standards#

Page 3Supko Greenberg §178.337-319) Traurig Caryo Tanks To. 2025313102 FOSTERA@gtlaw.com Fax 202.261.0102 07-0169 July 31, 2007 VIA COURIER 120 a ray apronion Washington, D.C. 20590 ¿ ALBANY Re: Meeting between Trinity Industries & DOT, July 10, 2007 AMSTERDAM ATLANTA Dear Ms. Gorsky: BOCA RATON We appreciate the opportunity to meet with you and your group and to express to you BOSTON our concerns over any Notice which might issue to the industry concerning the substance of BRUSSELS• Mr. Gale's letter of May 2, 2006, Ref. No. 06-0046 (the "Interpretation"). CHICAGO DALLAS, As we discussed at our meeting on July 10, 2007, we submit the following written DELAWARE comments for your consideration: DENVER FORT LAUDERDALE 1. This matter is not a safety issue. Trinity Industries, Inc. ("Trinity") began HOUSTON the 1950's. Trinity discontinued manufacturing the complete CTMVs in the mid-1970's and, manufacturing Cargo Tank Motor Vehicles ("CTMVs") and marketing them to the industry in LASVEGAS LONDON- LOS ANCiELES its history in this industry segment, it has manufactured at least two thousand CTVs or TCs since that time, has manufactured the Cargo Tanks ("CTs") only. Trinity estimates that, during MIAMI comparable to the model involved in the Flying J incident which prompted the instant MILAN- NEW JERSEY NEW YORK During its entire manufacturing experience (until the DOT advisory in May, 2006), CRANCE COUNTY Trinity manufactured the CTs (whether as a separate product or as part of a CTV) in the same | CRLANDO manner, to wit: by welding the internal baffle support clips directly to the inside surface of the PHILADELPHIA shell without the use of pads. Since the DOT advisory, Trinity has (while disagreeing with the PHOENIX need installed the clips using pads.' Approximately, 400 tanks have been manufactured with ROME". pads supporting the clips. SACRAMENTO SILICON VALLEY TALLAHASSCE TAMPA ICKYO• TYSONS CORNER 'This change by Trinity resolves the issue for its current and future CTs and customers. Further, at the present party. As a result, Trinity does not know which TCs have baffles and which do not. time, Trinity merely installs the clips on the pads. If the customer desires baffles, they are installed by a third WASHINGTON, D.C. WEST PALM BEACH ZURICH AL 76345044v1 7/31/2007 Greenberg Traurig, LLP | Attorneys at Law | 800 Connecticut Avenue, NW | Suite 500 | Washington, D.C. 20006 Tel 202.331.3100 | Fax 202.331.3101 www.gtlaw.com#

Page 4Page 2 of 3 July 31, 2007 During this experience of thousands of cargo tanks in service during 50 years of material or cracking of welds resulting from or related to the attachment of the baffle support production, Trinity has no knowledge, directly or indirectly, if any cracking of the shell clip directly to the shell without pads. company, ADAPCO, to determine the loadings on the baffle clips. ADAPCO measured the In addition, Trinity contracted with an independent engineering and consulting loads experienced in a sudden braking event and Trinity submitted them to ATECH Engineering to determine whether these loads posed a potential for failure. ATECH determined that even loads significantly higher were less than the maximum values allowed by the ASME code or by 49CFR 178.337. The ATECH certification is attached. kind to determine the cause of the crack in the shell wall of the CT. The crack could have been We would also point out that, as to the Flying J incident, there was no testing of any caused by one or more reasons, including the welding process, contaminated product inside the CT, or a combination of these or other reasons. Given the fact that the CT was only a few months old, fatigue seems highly unlikely. 2. Thus, this is a regulatory compliance issue. In that connection, Trinity does not believe that the current regulations govern, or even contemplate, the attachment of baffles to the interior surface of CTs. First, all examples of attachments in 178.337-3(g) are external in nature; indeed, this regulation makes no reference to any internal attachments at all. to the inside of the tank (because the inside radius of the pad cannot be determined by reference Furthermore, the design criteria for pads in §178.337-3(g)(3) (ii) cannot apply to a pad attached to the outside radius of the tank and, if it could, it could not equal the outside radius of the tank). Third, even if the provisions of §178.337-3 were ambiguous in whether they applied to attachments on the inside of the CTs, §178.337-5, which specifically address baffles, bulkheads requirement." The applicable rule of statutory and regulatory construction is that the specific and ring stiffness (all features on the inside of CTs), provides that they are "not a specification prevails over the general and, as result, there is no requirement concerning baffles. All these factors are consistent with the conclusion expressed in the email dated the baffles on the inside of the CTs, it would have to be amended [our copy of the email does February 21, 2006, from Mr. Shelton to Mr. Solomey, that, for the current regulation to cover not include the text of the entire "Recommended Regulations Text"]. DAL 76345044v1 7/31/2007 Greenberg Traurig, LLP#

Page 5July 31, 2007 Page 3 of 3 Interpretation of May 2, 2006.? In particular, Trinity requests that the Interpretation be 3. Requested action. Trinity respectfully requests that the DOT reconsider the rescinded on the grounds that it is not supported by the existing regulations and is causing cargo tanks to be removed from service without appropriate authority to support it.3 Further, the operating history of Trinity's MCTVs and CTs indicate that there is no immediate safety concern necessitating such immediate action. To the extent that the Department ultimately deems that, despite the absence of any history of safety concerns, internal pads should be utilized in connection with the installation of baffles, that conclusion should be expressed in an amended regulation of prospective application, established through the formal notice of rule making process, including a comment period. Again, thank you for the opportunity to meet with you and to share our views. Yours very truly, Allen - Poste C. Allen Foster Representing Trinity Industries, Inc. CAF:Icd cc: James O. Simmons Edward Muzzullo " In this connection, Trinity also respectfully points out that the instant discussions and exchange of views were initiated by Trinity in its continuing cooperative relationship with DOT. 3 Trinity estimates that the cost to the TC owner of removing existing clips, installing pads and reinstalling clips demonstrated safety concern is unreasonable. would run between $5.000-10.000 per TC. Trinity submits that to impose such costs on the industry without a DAL 76345044v1 7/31/2007 Greenberg Traurig, LLP#

Page 6U.S Department Of Transportation Memorandum Pipeline and Hazardous Materials MAY Safety Administration 2 2006 Ref. No. 06-0046 Subject: Cargo Tank Motor Vehicle From Johat A! Gale / Chief, Standards Development Office of Hazardous Materials Standards, OHMS Danny Shelton Federal Motor Carrier Safety Administration Hazardous Materials Division Regulations (HMR; 49 CFR Parts 171-180) for specification MC 331 cargo tank motor vehicles. This is in response to your letter concerning requirements in the Hazardous Materials 3 an anthe de it coin 178.32346, This ratament and etectives for September 1, 1995. For MC 331 cargo tanks manufactured after September 1, 1995, § 178.337- device to conform to the following requirements: 3(g) requires an attachment, appurtenance, structural support member, or accident protection 1. Appurtenances and other accessories must be attached to structural members, the reinforcement devices, when practicable. suspension sub-frame, accident protection structures, or external circumferential 2. A lightweight attachment to the cargo tank wall such as a conduit clip, brake line clip, percent of the thickness of the material to which it is attached. The lightweight having lesser strength than the cargo tank wall materials and may not be more than 72 3. Except as indicated above the welding of any appurtenance to the inside or outside of be no adverse effect upon the lading retention integrity of the cargo tank if any forc the cargo tank wall must be made by attachment of a mounting pad so that there wil less than that prescribed in § 178.337-3(b)(1) is applied from any direction. The#

Page 7thickness of the mounting pad may not be less than that of the shell wall or head wall must be drilled or punched at the lowest point before it is welded to the tank. date of manufacture. change, the design must be approved by a Design Certifying Engineer, and a Registered Inspector must certify that the cargo tank has been repaired and tested in accordance with the applicable specification. Section 180.413 specifies that repair work on an MC 331 cargo tank must be performed in accordance with the National Board Inspection Code (NBIC) and the Compressed Gas I hope this information is helpful. Please contact us if you require additional assistance ##

Page 8Greg McRae Engineering and Technical Director Thursday, July 19, 2007 2525 Stemmons Freeway Trinity Containers LLC 214 589 8559 Dallas, TX 75207 phone greg.mcrae@trin.net email Dear Mr. McRae Having modeled and analyzed a baffle clip configuration of the type used by Trinity Industries for the purpose of determining whether the cargo tank was adequate to withstand baffle clip loads having an estimated value of 155 Ib directed longitudinally and parallel with the tank axis, I have determined that both the highest concentrated principal stress values as well as the average principal stress values • in the cargo tank shell adjacent to the clip attachment; and • in the deposited fillet weld filler material; and • in the baffle clip at point of highest stress concentration, are significantly less than the maximum values allowed by the ASME code or by 49 CFR 178.337 for materials having the described mechanical properties. The model I used for analysis is that of a baffle clip connected by fillet weld attachment to the inside of a ¼ inch thick MC331 cargo tank vessel shell. The material properties I used in the analysis were consistent with properties of SA-517 quenched and tempered steel plate for the baffle clip construction and the tank wall, and with E11018 weld rod material for the fillet weld used to attach the clip to the tank wall. At the time of my analysis, I believed that the maximum load value to which any baffle clip would be subjected would not exceed the estimated value of 155 Ib. However, after having learned that the value actually determined from liquid surge analysis was only 135.1 Ibs, it is apparent that my conclusion regarding the effects of surge load on the baffle attachment configuration described is conservative. Ed Mansell DOT Registered DCE, CT # 8680 ATECH Engineering cc, Tom Rogers, Container Technology Inc.#