Full opinion text
MEMORANDUM AND ORDER HENDERSON, District Judge. This matter came before the Court on May 1, 2000, on plaintiffs’ and defendants’ cross-motions for summary judgment. Having carefully considered the written and oral arguments presented by the parties, intervenors, and amicus curiae, and the voluminous record herein, the Court grants in part, and denies in part, the motions for summary judgment as set forth below. FACTUAL BACKGROUND As explained in our December 10, 1997 memorandum addressing plaintiffs’ motion for preliminary injunction, this case has its genesis in two major railway accidents that occurred in 1991. On July 14, 1991, a train operated by the Southern Pacific Transportation Company derailed on the Can-tara Loop Bridge near Dunsmuir, California with devastating results. Thousands of gallons of metam-sodium, which becomes highly toxic when mixed with water, spilled into the Sacramento River, killing all fish and vegetation along the river for 40 miles, and causing widespread health problems for area residents. The Sacramento River ecosystem may never recover given that some species were completely extricated from the river. See Bagley Deck, Vol. 1, ¶ 2, and appended Administrative Record (hereafter “AR”) 2656-57. Two weeks later, on July 28, 1991, a Southern Pacific train was involved in another derailment near Seacliff, California. The release of toxic materials led to evacuations, the closing of a highway for approximately five days, and medical treatment for many. Id. at 2657; 57 CPUC2d 368. The California legislature responded to these and other railroad accidents by directing the California Public Utilities Commission (“CPUC”) to identify “local safety hazards” on California railways and to adopt regulations “to reduce the potential railroad hazards” at those sites. See Cal. Pub.Util.Code §§ 7711(d), 7712; see also id. at § 765.5 (1991) (amended 1999); AR 2644-45. In August of 1991, the CPUC issued an order instituting an investigation into the July 14, 1991 Cantara Loop Bridge derailment. CPUC concluded that the derailment was caused by track-train dynamics, and in particular the eonfiguration of the train cars, which in this case included heavy/light car and short/long car combinations too close to the head-end of the train. Re Southern Pacific Trans. Co., 57 CPUC2d 386, 400-01, 1994 WL 746871. (1994). In December of 1994, the CPUC issued its decision regarding the Dunsmuir investigation, which found the Cantara Curve to be a local safety hazard. The CPUC issued rules regulating trains traveling over that site, including (1) a requirement that Southern Pacific utilize a state-approved locomotive maintenance quality improvement program; and (2) a requirement that trains operating over -the Cantara site be configured in a specific manner. See Decision No. 94-11-069 at 403-06 (hereinafter “Dunsmuir Decision”). Meanwhile, in March of 1992, the CPUC issued an order instituting an investigation into potential railroad safety hazards in California generally. See Order Instituting Investigation No. 92-03-17. AR 12-44. On September 3, 1997, the CPUC issued a 174-page decision, 97-09-045, which identified 19 sites, located in and around California mountains, as “local safety hazard sites” and adopted numerous regulations governing operations at these sites. AR 2640-2817 (hereafter “Decision”). Among other things, the CPUC’s order required the railroads to: (1) cooperate in developing performance-based standards for train, configuration based on track-train dynamics principles; (2) develop standards for dynamic braking systems; (3) equip trains with two-way end-of-train telemetry devices; (4) institute new training programs regarding track-train dynamics principles; (5) install trackside defect detectors at 30 mile intervals; (6) adopt heightened standards for securing standing trains; (7) maintain current track strength at one particular site; and (8) not discipline railroad employees who report violations of the new regulations (the “whistleblower rule”). The Decision explained that: [W]e implement these regulations not out of any sense of competition or dissatisfaction with the [Federal Railroad Administration (“FRA”) ] but, rather, out of sheer necessity to protect California’s people, its environment and its commerce against the disastrous consequences of recent rail accidents and toxic spills. In issuing this Decision, we intend to complement the FRA’s efforts and hope that both the Railroads and the FRA will join us in securing greater safety and fewer accidents in railroad operations in this state.” Id. at 2. On October 9, 1997, plaintiffs Union Pacific Railroad (“Union Pacific”), Southern Pacific Transportation Company (“Southern Pacific”), and Burlington Northern and Santa Fe Railway Company (“Burlington Northern”) sued to enjoin these CPUC regulations as preempted by (1) the Federal Railroad Safety Act, 49 U.S.C. § 20101 et seq.; (2) the Locomotive Boiler Inspection Act, 49 U.S.C. § 20701 et seq.; (3) the Safety Appliance Act, 49 U.S.C. § 20301 et seq.; (4) the Hazardous Materials Transportation Act, 49 U.S.C. § 5125; and (5) the Commerce Clause of the United States Constitution. Plaintiffs also claimed that aspects of two of the rules violate the First Amendment. Finally, plaintiffs sought to enjoin three state statutes governing railroad safety issues, Calif.Pub.Util.Code §§ 7672(b) — (c), 7672.5, 7673(c), as preempted by the Hazardous Materials Transportation Act, 49 U.S.C. §§ 5101 et seq., and the Federal Railroad Safety Act, 49 U.S.C. § 20101 et seq. Several environmental groups (“the Cantara intervenors”) and unions (“the union intervenors”) intervened as party defendants. On November 26, 1997, this Court granted in part and denied in part plaintiffs’ motion for preliminary injunction. On December 14, 1998, this Court granted partial summary judgment on some of the preemption issues presented by this action, holding that (1) Calif.Pub.Util.Code § 7672.5 is not preempted by the Hazardous Materials Transportation Act (“HMTA”) or the Federal Railroad Safety Act, and (2) that Calif.Pub.Util.Code § 7673(c) is preempted by the HMTA. It also reconfirmed an earlier ruling that Calif. Public UtiLCode § 7672(b)-(c) is preempted by the HMTA. Now the parties move for summary judgment on all but two remaining claims. The Cantara Intervenors have joined in the CPUC’s opposition and cross-motion for summary judgment. The Union Inter-venors have filed separate papers in support of the CPUC. The United States of America, on behalf of the U.S. Department of Transportation (“DOT”) and the Federal Railroad Administration (“FRA”), and the Association of American Railroads have also submitted briefs as amicus curi-£16. DISCUSSION I. PREEMPTION BY FEDERAL STATUTE OR REGULATION Under the Supremacy Clause, Congress may supplant state regulation with uniform national rules. Law v. General Motors, Corp., 114 F.3d 908, 909 (1997). “Given the importance of federalism in our constitutional structure, however, we entertain a strong presumption that federal statutes do not preempt state laws; particularly those laws directed at subjects — like health and safety — ‘traditionally governed’ by the states.” Id. at 909-10. Accordingly, a finding of preemption must rest on the clear and manifest purpose of Congress. Id. at 910. Congress has addressed the issue of railroad safety and equipment in the three statutes at issue in these motions-— the Federal Railroad Safety Act (“FRSA”), the Locomotive Boiler Inspection Act (“LBIA”), and the Safety Appliance Act (“SAA”) — and delegated the administration of these laws to the Secretary of Transportation, who in turn has delegated implementation authority to the FRA, an operating administration within DOT. See 49 U.S.C. § 20103; 45 U.S.C. § 431; 49 C.F.R. § 1.49(n); CSX Transp. v. Easterwood, 507 U.S. 658, 662, 113 S.Ct. 1732, 123 L.Ed.2d 387 (1993). Together, these laws and their implementing regulations create an extensive network of federal railroad regulations; nonetheless, Congress has not occupied the entire field of railway regulation, and states may still promulgate and independently enforce rules in certain circumstances. 49 U.S.C. § 20106; See also Bagley Decl., Part II, Exh. E, App. B. at 187 (FRA Report to Congress acknowledging that while uniformity is emphasized by the FRSA, “state regulatory action can often point the way for national standards.”) In determining the preemptive effect of the FRSA, the LBIA, and the SAA, the Court is, of course, bound by their terms. “First, always, is the question of whether Congress has directly spoken to the precise question at issue. If the intent of Congress is clear, that is the end of the matter; for the court, as well as the agency, must give effect to the unambiguously expressed intent of Congress.” Chevron v. Natural Resources Defense Council, 467 U.S. 837, 842-43, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984). Where, however, the statute is silent or ambiguous, then courts must defer to the agency’s permissible construction of the statute. Id. at 843, 104 S.Ct. 2778. In particular, agency regulations that address a gap left in the statute, or ambiguous provisions, must be given controlling weight unless they are arbitrary, capricious or manifestly contrary to the statute. Id. at 844, 104 S.Ct. 2778. As to the preemptive effect of regulations promulgated under the FRSA, LBIA or SAA, it is established that “[a]n agency’s “interpretation of the preemptive effect of its regulations is entitled to deference where Congress has delegated authority to the agency, the agency’s interpretation is not contrary to a statute, and agency expertise is important to determining preemption.” Industrial Truck Assoc., Inc. v. Henry, 125 F.3d 1305, 1311 (9th Cir.1997). With these general principles in mind, the Court turns to the specific statutes and regulations at issue in this case. A. THE FEDERAL RAILROAD SAFETY ACT The most recent Congressional enactment in the area of railroad operations— the Federal Railroad Safety Act (“FRSA”), 49 U.S.C. § 20101 et seq. — was adopted in 1970 as a supplement to the Safety Appliance Act and the Locomotive Boiler Inspection Acts which were passed in 1915 and 1911 respectively. Explicitly addressed to the issue of safety, the FRSA’s stated purpose is “to promote safety in every area of railroad operations and reduce railroad-related accidents and incidents.” 49 U.S.C. § 20101; CSX, 507 U.S. at 661, 113 S.Ct. 1732; Southern Pac. Transp. Co., 9 F.3d 807, 812 (9th Cir.1993) (FRSA was enacted “to reduce railroad-related accidents and deaths and to improve rail safety more generally”). It also declares that the rules and laws regulating railroad safety should be “nationally uniform to the extent practicable.” 49 U.S.C. § 20106. Congress balanced these objectives by expressly reserving a role for state regulation in the area of railroad safety in two circumstances. First, a state may adopt any railroad safety rule if there is no federal regulation “covering the subject matter of the state requirement.” Second, even where federal regulation already “covers” the subject matter, a state may enforce a more stringent rule if it can demonstrate that it is necessary to address a local safety hazard, is not incompatible with federal law, and will not unreasonably burden interstate commerce: A State may adopt or continue in force a law, regulation, or order related to railroad safety until the Secretary of Transportation prescribes a regulation or issues an order covering the subject matter of the State requirement. A State may adopt or continue in force an additional or more stringent law, regulation, or order related to railroad safety when the law, regulation, or order— (1)is necessary to eliminate or reduce an essentially local safety hazard; (2) is not incompatible with a law, regulation, or order of the United States Government; and (3) does not unreasonably burden interstate commerce. 49 U.S.C. § 20106. In short, “[u]nder [the FRSA] ... scheme ... state regulations can fill gaps where the Secretary has not yet regulated, and it can respond to safety concerns of a local rather than national character.” Burlington N. and Santa Fe Ry. Co. v. Doyle, 186 F.3d 790, 795 (7th Cir.1999). Thus, in order to prevail on their claim of FRSA preemption, plaintiffs must establish both (1) that the “subject matter” of the challenged rules has already been “covered” by the FRSA, and (2) that the “local safety hazard” exception does not apply. 1. Whether the FRSA “covers” the “subject matter” of the CPUC rules It is well established that FRSA regulations do not “cover” the subject matter of a state rule by simply “touching upon” or “relating to” the subject matter; rather, “cover” is a more restrictive term that only applies if the federal regulations substantially subsume the subject matter of the relevant state law. CSX, 507 U.S. at 664, 113 S.Ct. 1732 (emphasis added); see also Norfolk Southern Railway Co. v. Shanklin, 529 U.S. 344, 120 S.Ct. 1467, 1473, 146 L.Ed.2d 374 (2000); Southern Pac. Transp. Co., 9 F.3d at 812. Indeed, the Ninth Circuit has observed that the term “covering” is employed within a provision that “displays considerable solicitude for state law,” and notes that FRSA preemption is “even more disfavored than preemption generally.” Id. With respect to characterizing the “subject matter” of the state rule, courts must look to the safety concerns that the state law addresses. Burlington Northern R.R. v. Montana, 880 F.2d 1104, 1106 (9th Cir.1989) (“[The FRSA] preempts all state regulations aimed at the same safety concerns addressed by FRA regulations”). As courts have acknowledged, if the subject matter of the state rule is too narrowly defined, a state law could only be preempted if there were an identical federal regulation, which is not what Congress intended. See Burlington Northern, 186 F.3d at 796. On the other hand, an overly broad characterization of the subject matter— e.g. “rail safety” — would render the analysis meaningless because all FRA regulations cover this concern. See id. Each rule challenged under the FRSA is addressed in turn below. (а) CPUC Rules on Track — Train Dynamics Train-track dynamics (“TTD”) is a broad area covering almost any subject that affects a train’s ability to remain on the tracks (aside from a collision). See King Decl. at ¶ 8. It generally refers to the interplay of forces between a train and the track on which it is traveling, and covers a broad range of issues from track geometry to speed limits and train handling techniques. The CPUC TTD rule is focused on one element of TTD — the issue of train configuration or “make-up.” As the DOT explains, train configuration or “make-up” focuses on the order in which a train is assembled, “with consideration for the relative weight and dimensions of the various vehicles that make it up in order to facilitate proper train handling when braking, on heavy grades, and on curves ...” DOT brief at 9. Specifically, Train make-up ... involves placing cars in a train such that they balance the forces within the train. Here, relevant considerations include empty versus loaded cars, short versus long cars, and the effects of terrain and curvature. Bagley Decl., Part II, Exh. E at 67 (FRA report to Congress). There are currently no federal train make-up rules. Rather, train make-up is governed by the railroads’ own operating rules. The CPUC Decision imposes three discrete TTD train make-up rules. First, it requires the railroads to cooperate with CPUC staff and others in developing and implementing performance-based train make-up standards for 13 identified local safety hazard sites. Second, it requires the railroads to comply with their own train make-up rules for these sites as set forth in their internal operating rules. Third, it orders the railroads to obtain CPUC approval before making any changes to these rules. See Decision at 168-70, ¶¶ 1-11; App. A3-A4; King Deck at 3-5. Plaintiffs argue that all three rules are preempted by the FRSA. (i) CPUC rule requiring Railroads to cooperate in developing and implementing train make-up standards for IS sites Plaintiffs argue that FRA rules already cover the subject matter of TTD through a variety of related measures that address various aspects of TTD, such as train speed and Lateral to Vertical (L/V) force ratios. As DOT makes abundantly clear, however, “the FRA has not issued a rule or order on train make-up” — the specific aspect of TTD at issue here. DOT brief at 14; see also id. at 13, n. 7 (“No such federal standard exists with regard to track-train dynamics”); id. at 12 (“the state is free to regulate track-train dynamics by establishing state standards [consistent with any commerce clause constraints]”); Elsmore Decl. at ¶ 5, p. 4 and Exh. 3 (FRA’s Operating Practices Compliance Manual) (“NOTE: There are no Federal regulations governing train make-up, except for train placement of certain hazardous materials cars.”); Elsmore Suppl.Decl. ¶¶ 6-8; Bag-ley Decl. Part II, Exh. E at 67 (FRA report to Congress noting that FRA ivill launch formal regulatory action in the area of train make-up in 1996). As of yet, however, no such rulemaking has been initiated. See King Decl. at ¶ 12. Plaintiffs also argue that the FRA has already explicitly considered and rejected the notion of mandating specific train make-up rules. They emphasize that in 1982, the National Transportation Safety Board (“NTSB”) recommended that FRA adopt track-train dynamics rules but the FRA declined. See 47 Fed.Reg. at 32,818 (stating that FRA “does not plan to issue regulations pertaining to train make-up or train handling.”). This affirmative decision not to regulate, they contend, preempts any state action in the area of TTD. While it is certainly true that FRA’s “rejection, like its adoption, of particular safety regulations may preempt state regulations on the same subject matter,” Burlington N.R.R., 880 F.2d at 1107, this Court agrees with DOT that “FRA has not taken any action that would amount to negative preemption in this area.” DOT brief at 14. First, as DOT notes, the NTSB reeommen-dations pertained to trains carrying hazardous materials. See DOT brief at 14 (NTSB recommendations were made “in the hazardous materials context”); 43 Fed. Reg. 35,563, 35,564, 59,557, 59,558 (1978) (discussing issue in context of cars carrying hazardous materials). Second, as DOT explains: FRA’s decision not to adopt train makeup rules in 1982 did not constitute an affirmative determination that such regulation was not needed and that relevant requirements should not be imposed by the States. Instead, FRA concluded that it did not then know how to regulate train make-up effectively and that safety would be better served through other actions FRA could and did take. DOT brief at 14-15. Because the FRA has not substantially subsumed, and thus has not covered, the area of train make-up, the FRSA does not preempt CPUC’s order that the railroads cooperate with CPUC staff and others to develop performance-based standards for train configurations based on current track-train dynamics principles. Indeed, plaintiffs have fallen far short of establishing that it would be impossible to develop train make-up standards that would not be covered by the FRSA. Compare infra section 1(B)(2) (finding that CPUC rule requiring plaintiffs to help develop dynamic brake rule is preempted because any such rule would necessarily be preempted). (ii) CPUC rule requiring CPUC approval of changes to railroads’ train make-up operating rules at the IS sites In contrast to the subject of train makeup, which the FRA has yet to regulate, the FRA has acted in the area of railroad operating rules. Specifically, 49 C.F.R. §§ 217.7-11 requires that the railroads (1) file copies of their operating rules with the FRA, (2) conduct their own self-monitoring tests and inspections to determine compliance with their operating rules, (3) keep records of and report annually on those self-monitoring tests and inspections, and (4) periodically train their employees on the operating rules. FRA regulations do not, however, mandate the content of the rules (except for a few very limited exceptions, see e.g. 49 C.F.R. §§ 218-219; Els-more Decl. at ¶ 24); nor do they require that the rules be approved by the FRA. By virtue of the fact that the FRA does not, with minor exception, regulate the content of the railroads’ internal operating rules, the FRA has effectively deferred to the railroads with respect to the content of the rules (subject of course to any governing federal regulation). This deference in turn necessarily implies that FRA has also determined that no approval of railroad operating rules is warranted. Indeed it would be totally inconsistent to defer to the railroads regarding the content of their rules, but then require that the rules, or modifications thereto, be subject to “approval” by the different states. Given the above, this Court concludes that the FRA’s regulations governing the railroads’ operating rules, set forth in 49 C.F.R. § 217 et seq., “cover” the issue of approval of the railroad’s operating rules. (Hi) CPUC rule requiring that railroads comply with their own operating rules, with respect to train make-up at the IS sites As noted above, the Decision also requires that the railroads comply with them own train make-up rules (as set forth in their own operating rules and related documents) at 13 sites. See Decision at 169, ¶ 8. While DOT acknowledges that this rule presents a closer question, it contends that 49 C.F.R. § 217 et seq., described above, does not cover the subject matter of whether railroads may be “required” to abide by their own operating rules. Rather, it submits, 49 C.F.R. § 217 et seq. covers only the more narrow subject of “self-monitoring” by railroads to determine the extent of their compliance with their own rules. DOT brief at 13-14. Under this view, FRA has not covered the issue of whether railroads can be required to comply with their own operating rules. Plaintiffs emphasize that the safety concern of the CPUC’s rule is to promote better train make-up through ensuring compliance with railroads’ own rules. They then argue that because the self-monitoring provisions of section 217 are aimed at this same safety purpose of “compliance,” it covers the subject of the CPUC rule. While this argument has appeal at first blush, it does not survive closer scrutiny. While the “self-monitoring” requirements in § 217 certainly touch upon and relate to the issue of compliance, it does not appear that the focus of the self-monitoring requirements is to ensure that the railroads are complying with their own train makeup rules. Rather, FRA explained that it was requiring the railroads to conduct, and report on, the self-monitoring tests and inspections because “[tjhe information to be gained through the implementation of these requirements is considered necessary to the formulation of uniform operating rules.” 38 Fed.Reg. 12617, 12618 (May 14, 1973); see also Elsmore Decl. ¶ 28. Indeed, the fact that the FRA takes no “compliance-related” action with respect to the self- monitoring information, further suggests that the purpose of the self-monitoring provisions is not to ensure compli-anee. While reasonable minds could differ on this issue, the FRA’s view of the preemptive scope of its regulation is not an unduly narrow or unreasonable one and thus deserves deference. Industrial Truck Assoc., 125 F.3d at 1311. Moreover, the Court is mindful that the “covering” standard is a “stringent” one, and of the general “presumption against pre-emption” in this area. CSX, 507 U.S. at 668, 113 S.Ct. 1732; Southern Pac. 9 F.3d at 812. Accordingly, this Court concludes that the FRA regulations do not substantially subsume or cover CPUC’s rule requiring the railroads to comply with their own train make-up rules at 13 sites. In sum, this Court finds, with respect to the train make-up rules, that (1) CPUC’s order requiring the railroads to cooperate in the development of train make-up rules is not covered by the FRSA, (2) CPUC’s order requiring approval of changes to the railroads’ operating rules is covered by the FRSA, and (3) CPUC’s order requiring the railroads to comply with their own train make-up rules is not covered by the FRSA. (b) CPUC Trairir-Track Dynamics (TTD) Training Rule The CPUC’s Decision requires that train crews undertake additional training regarding the make-up rules for six sites which pose particularly difficult TTD-related hazards and are subject to specialized train make-up rules. See Decision at 171-72 and App. A-4; King Decl. at ¶¶ 122-23 (requiring additional instruction, testing, and periodic certification of employees with respect to the railroads’ train makeup rules for certain sites). As CPUC concedes, the issue of training is addressed by the FRA in 49 C.F.R. §§ 217.11, 240.123, 213. These regulations require that railroads provide initial and periodic instruction for their employees on the railroad’s operating rules, to “ensure” the employees’ “understand[ing]” of and “meaning and application” of the rules. Id. at § 217.11. They also provide standards for the initial and continuing education for locomotive engineers, including education on train handling, operating rules and the physical characteristics of the railroad. CPUC contends, however, that the FRA regulations are overly general and fail to require needed site-specific training. See e.g. Elsmore Decl. ¶ 3, et seq. While defendants may be dissatisfied with the rules, and consider them weak and ineffectual, the question before the Court is not whether the FRA rules could stand improvement, but whether they “substantially subsume” the subject of the state rule, which is defined by the safety concern that it addresses. Given that the safety concern of the CPUC’s training rule can be fairly described- — -not too narrowly and not too broadly — as “training for hazardous sites,” the Court concludes that the FRA has covered the subject. The FRA specifically requires training on all operating rules, which plaintiffs credibly represent to include training on all General Orders, Superintendent’s Notices, timetables and special instructions. See Patón Decl. ¶¶ 7-11, 23; Patón Suppl. Decl. ¶ 15; Wills Decl. ¶ 37, Repola Decl. ¶ 10, and Exh. A. As such, the FRA already requires training on the railroads’ train make-up rules. Id.; see also Quilty Decl. ¶ 9 (BNSF tests on all rules including general orders and system special instructions which address make-up issues); Patón Supp.Decl. at ¶¶ 14-17; Wills Decl. ¶ 25, and Exh. B; 49 C.F.R. § 217.11(a). Notably, the FRA has also expressly delegated the training function to the railroads, subject to intervention by the FRA and state inspectors. See 49 C.F.R. § 217.11 (“[E]ach railroad to which this part applies shall periodically instruct each such employee on the meaning and application of the railroad’s operating rules ... ”) (emphasis added); Patón SuppLDecl. at ¶ 9.-12. Accordingly, and consistent with the view of the DOT, this Court concludes that the CPUC’s training regulation concerns a subject that is already “covered” by federal regulation. (c) CPUC Rule on Trackside Defect Detectors (“Hot Box” Detectors) Trackside defect detectors (also referred to as “hot box” detectors) are devices attached to railroad tracks that monitor whether passing trains are operating with overheated wheel bearings, a condition that makes train derailment more likely. If the monitor senses that the bearing is overheated, it sends a signal to a dispatcher, wayside display, or the train crew. The CPUC’s Decision requires the railroads to install hotbox detectors at intervals of no greater than 30 miles. See Decision at 172, ¶ 23. Because the railroads generally place wayside detectors roughly every 30 miles, as a practical matter, this rule only requires Union Pacific to install a single detector near site No. 25 (where no detector is in place for over 60 miles), because all other locations are already in compliance with the 30-mile rule. See Decision at 93. While plaintiffs argue that a number of federal regulations already “cover” the subject area of hot-box detectors, in fact they only touch upon, rather than substantially subsume, the subject. Plaintiffs, for example, argue that the FRA specifically rejected proposals to require hot-box devices in 58 Fed.Reg. 36,-605, 36,607 (1993). There, the FRA issued a final rule requiring event recorders on fast trains that can provide quality information for post-accident investigations. In issuing this rule, the FRA decided that it was not necessary to have hot-box detector data recorded by the event recorders because information regarding hot box detector readings is already available through other sources. See id. at 36,607. At best, this FRA rule only “touches” upon the subject as it relates to the issue of event recorders. It certainly does not substantially subsume the question of where hotboxes should be installed. Plaintiffs’ reliance on 49 C.F.R. § 215.115 is similarly misplaced. That section prohibits railroads from using trains with overheated or defective bearings (which hot-box detectors are intended to identify) and requires that bearings on cars that have been involved in derailments be physically inspected for defects. Thus, while section 215.115 addresses the hazard that results from defective bearings, it does not speak to the use of proactive measures, such as hot-box detectors, to detect defective bearings before they become hazardous. Nor does 49 C.F.R. § 238.427(e) assist plaintiffs since it concerns passenger, not freight, trains. Finally, plaintiffs’ citation to a FRA Report to Congress under the Hazardous Materials Act is also inapposite. Accordingly, this Court agrees with the DOT that plaintiffs have failed to demonstrate that the subject of trackside defect detectors has been “covered” by the FRSA. As such, this rule is not preempted by the FRSA. (d) CPUC Rule on Track Standards In 1983, SP upgraded, considerably above federal minimum levels, the strength of its track running across a ten-mile track segment in Site No. 9, which includes the Cantara Loop (site of the 1991 accident near Dunsmuir). The Decision requires Southern Pacific to maintain the track strength at above minimum federal standards and to obtain CPUC approval for any change. See Decision at 172-73, ¶¶ 24-27. It is clear that federal regulations substantially subsume the subject of track safety standards. As the DOT aptly describes it, 49 C.F.R. Part 213 sets forth “comprehensive” federal track safety standards. See DOT brief at 16. Notably, these standards, recently amended in 1998, specifically address track strength at curves. See 63 Fed.Reg. 33,992, 34,006; see also 49 C.F.R. § 213.109. In its response to DOT’s brief, defendants also concede “that track standards are ‘covered’ by FRA rules.” Defendants’ response to DOT amicus at 13. Plainly, CPUC’s track standard rule is covered by the FRSA. (e) CPUC Rule on End-of-Train Telemetry Device End-of-train telemetry (“EOT”) devices monitor and regulate the air brake system on a train. As discussed below this Court concludes that CPUC’s rule regarding EOT devices is preempted by the LBIA. As such, it is unnecessary for the Court to also address whether the EOT rule is “covered” by the FRSA, and if so, whether it is nonetheless saved from preemption by the local safety hazard exception. Accordingly, it declines to do so. (f) Summary As set forth above, this Court finds that two of the three train make-up rules are not “covered” by the FRSA: (1) CPUC’s rule requiring the railroads to cooperate in the development of train make-up rules and (2) CPUC’s rule requiring the railroads to comply with their own train make-up rules. In addition, CPUC’s rule on trackside defect detectors (the “Hotbox” rule) is not covered by the FRSA. As such, plaintiffs’ contention that these rules are preempted by the FRSA must be rejected. The Court finds, however, that the Secretary has “covered” the subject matter of the following three rules: (1) the rule requiring state approval of changes to the railroads’ TTD operating rules, (2) the employee TTD training rule, and (3) the track standards rule. As such, these rules are preempted by the FRSA unless defendants can establish that they are saved by the local safety hazard exception, which issue the Court turns to below. 2. Whether the “Covered” Rules are Permissible Under the “Local Safety Hazard Exception” As described above, even when the Secretary has already “covered” the subject area of the state regulation, Congress provided that a state may still enforce an “additional or more stringent law, regulation, or order related to railroad safety, when the law, regulation, or order (1) is necessary to eliminate or reduce an essentially local safety hazard; (2) is not incompatible with a law, regulation, or order of the United States Government; and (3) does not unreasonably burden interstate commerce.” 49 U.S.C. § 20106. As discussed below, this Court finds, with respect to the three rules that are covered, that defendants have not demonstrated that the rules requiring (1) additional training on make-up rules affecting six identified local safety hazard sites, and (2) state approval of changes to the railroad’s make-up rules, are “necessary” to eliminate or reduce an essentially local safety hazard, even assuming arguendo that the affected sites qualified as local safety hazards. Accordingly, these latter two rules are preempted by the FRSA. The Court does find, however, that the track standards rule, which affects local safety hazard site No. 9, is saved from preemption under the local safety hazard provision. (a) TTD Training Rule The Court need not determine whether the six affected sites constitute “essentially local safety hazards,” because even assuming arguendo this showing could be made, defendants have not established that the additional instruction, testing, and periodic certification requirements they seek to impose are necessary to reduce or eliminate any TTD hazard at those sites. First, defendants simply have not shown that there is a training problem to fix in that they have not demonstrated that railroad employees are in fact failing to follow the specific TTD rules for the sites at issue, or that deviation from those rules are causing derailments at the affected sites. Second, even assuming arguendo that the current federally-mandated railroad training programs are not satisfactorily covering the special TTD rules for the six sites, defendants have not satisfactorily shown that a separate state-imposed training program is necessary to address the problem. Rather, under the federal system in place, CPUC has the authority to inspect and monitor the railroad’s training programs. If problems are identified, it can seek corrective and other enforcement action. See 49 C.F.R. § 212.115; Patón Decl. ¶¶ 12-14; Patón SuppLDeel. at ¶ 11; Wills Decl. ¶¶ 41-43. CPUC admits, however, that it has never initiated any enforcement proceedings on the ground that training regarding train make-up is inadequate. See Bagley Decl., Exh. K at 13-14; Repola Decl. ¶ 13. Because defendants have never attempted to address their training concerns through the existing federal framework, they can not show that imposition of a supplemental state training program is necessary. Accordingly, CPUC’s TTD training rule is preempted by the FRSA. (b) Rule Requiring CPUC Approval of Changes to TTD Operating Rules As noted above, California has reserved to itself the power to veto or approve any changes in the TTD train make-up rules affecting 13 identified local safety hazard sites. See Decision at 169, ¶¶ 8-9 and Appendix A. Again, however, even assuming arguendo that the sites qualify as local safety hazards, California has not demonstrated that the rule is necessary to reduce any TTD-related hazards at these sites. As in the case of training programs, the federal framework already provides a mechanism by which states can oversee the actions of the railroads. Federal and state inspectors are authorized to review railroad operating procedures, including train make-up rules, identify any unsafe or inadequate procedure, and seek corrective action. See Patón Decl. ¶¶ 21-23; Patón Suppl.Decl. ¶¶ 16-17. While the parties may disagree as to the efficacy of this system, defendants have not documented deficiencies sufficient to justify finding that a separate state regulatory approval-scheme is necessary to reduce any TTD-related risks at the sites at issue. This is particularly so given that permitting states to exercise veto power over changes in railroad operating rules involves a substantial intrusion into the flexibility afforded railroads with respect to their operating rules under the federal scheme. Certainly, a substantially greater showing than that made here would be necessary to justify the overlay of state regulation promulgated by the CPUC. Accordingly, CPUC’s rule requiring state approval of any changes to the railroads’ train make-up rules for the affected sites is preempted by the FRSA. (c) Track Standards Rule As previously described, CPUC’s track standard rule requires plaintiffs to maintain the track standard at the Cantara Curve (site 9) at a level exceeding minimal federal standards. In contrast to the rules discussed above, defendants have shown, as explained below, that this rule is “necessary.” Accordingly, the Court must confront whether site 9 constitutes “an essentially local safety hazard” and whether the latter two prongs of the local safety hazard provision are met as well. (i) Essentially local safety hazard The question of what constitutes an “essentially local safety hazard” raises a number of difficult and sparsely litigated issues. Specifically, the Court must address (1) CPUC’s methodology for identifying local safety hazards, (2) the proper definition of an “essentially local safety hazard,” and (3) whether site 9 qualifies as a local safety hazard, as that term is properly defined. (A) Methodology for identifying an essentially local safety hazard As a threshold matter, plaintiffs and DOT argue that the CPUC’s methodology for identifying local safety hazard sites was so deficient that all of the 19 identified local safety hazard sites should be rejected on this ground alone, without the need for any individualized review. This Court concludes, however, that while the methodology has flaws, site 9 should be evaluated on its own merits. The CPUC began its process by using three independent, statistical methods (the Poisson, binomial, and multinominal models) to verify that accident concentrations at particular locations were not possibly random occurrences but could only be explained by site characteristics. Decision at 26-36. Plaintiffs have not disputed the validity of CPUC’s statistical calculations in this proceeding. CPUC then screened the statistically identified sites for accident cause codes that tied the nature of the hazard to the location. King 1st Am.Decl. ¶¶ 59-63. If the causes did not indicate a particular local problem, the sites were dropped from consideration. Decision at 123-143; King Suppl.Decl. ¶¶ 65, 69. CPUC also identified those mountain grade sites in which extreme conditions forced the railroads to follow specially tailored, self-imposed operating restrictions. Decision at 36. Most of these sites overlapped with the sites that were statistically identified. Id. CPUC also considered demographic and environmental factors, and other local features. All of the above factors were considered together to make an individualized assessment as to each site. Id. at 123-43: Thus, CPUC justified its ultimate decision to designate each of the 19 locations as local safety hazards by explaining the specific combination of factors that, in its view, renders each site to be a local safety hazard. See Decision at 120-143. No site was designated because of derailment frequency or operating restrictions alone. Given the above, plaintiffs’ and DOT’s contention that CPUC’s approach must be rejected because it identified local safety hazards based on accident clusters or operating restrictions alone is based on an oversimplification of CPUC’s approach. The argument that the sites are based on grade and curve alone is similarly misleading. See King Suppl.Decl. at ¶¶ 71-79. Rather, the record shows that CPUC engaged in and offered an analysis of various factors before designating any site to be a local safety hazard. DOT also argues that it was inappropriate for California to go out and “search for [local] safety problems across the entire state ...” DOT brief at 21. This Court, however, finds nothing in the FRSA that would prohibit states from taking a proactive approach to identifying genuine, local safety hazards within their borders. Maximizing railroad safety is, after all, the primary purpose of the FRSA, see 49 U.S.C. § 20101, and plainly the motivating factor behind the “local safety hazard” savings clause. While the Court is cognizant that to a man with a hammer, many things may look like nails, this concern can be addressed by careful scrutiny of the basis for any state designated local safety hazard. It does not, however, preclude states from making a state-wide assessment of railway lines within their borders to determine the existence, if any, of “essentially local safety hazards.” California’s methodology could certainly have benefitted from further refinement, and in many cases CPUC failed to analyze whether the site featured any local conditions or factors that were genuinely uncommon, making the designation of many of the sites questionable. Nonetheless, given all of the above, this Court declines to reject in blanket fashion all 19 of the designated local safety hazard sites. Rather, it concludes that sites should be scrutinized on an individual basis. Accordingly, the Court turns to the question of whether site 9 represents a local safety hazard. (B) Definition of essentially local safety hazard As noted above, an essentially local safety hazard is a safety hazard that is not statewide or national in character, but rather addresses a specific local concern. Doyle, 186 F.3d at 795. While the case law in this area is relatively sparse, courts have consistently and predictably held that state railroad safety rules that have general state-wide applicability can not survive under the local hazard exception. See e.g. Coleman, 542 F.2d at 14-15 (invalidating statewide accident reporting requirements because they were not directed at reducing local safety hazard); Consolidated Rail Corp. v. Pennsylvania Pub. Util. Comm’n, 536 F.Supp. 653, 658 (E.D.Pa.1982) (holding that state law requiring all locomotives to have speed recorders and indicators did not relate to “a local problem” and that state can not invoke local safety exception on ground that the state, itself, has a “unique statewide problem”), aff'd 696 F.2d 981 (3rd Cir.1982). Courts have also found that regulations, which on their face, do not apply statewide, but are still too geographically general in scope, will not satisfy the “local safety hazard” exception. For example, in Union Pacific Railroad Co. v. Public Util. Comm’n of Oregon, the State of Oregon imposed additional safety requirements on trains operating through local safety hazards, which were defined as 1) those areas sensitive to hazardous waste materials; 2) gradient areas, 3) urban areas; and 4) railroad-highway grade crossings which have maximum speed limitations. 723 F.Supp. 526, 528 (D.Or.1989). Under this definition, approximately 21 percent of train track in Oregon fell within one or more of the above four categories and thus ran through a local safety hazard. Id. (In contrast, all 19 of the California sites comprise 4.2% of all California track. See Decision at 5). The Court concluded that the hazards identified were “not peculiar” to either the State of Oregon or parts of the Oregon. Instead they “involve the types of terrain and crossing that commonly occur throughout the national railroad system,” and therefore “are for practical purposes statewide.” Union Pacific, 723 F.Supp. at 530. Similarly, in Burlington v. Montana, the district court found that a Montana law that required a two-way telemetry system on all “mountain grades” was not sufficiently local to qualify for the local hazard exception. 805 F.Supp. at 1522. “Mountain grades,” the Court observed, “occur in many places of Montana and throughout the nation and are not peculiar to a particular locality.” Id. at 1528. The Court was also concerned that the “mountain grade” sites were defined administratively (by reference to mile posts in the railroad’s timetable and operating rules), and not geographically. Id. These rulings are consistent with the FRSA’s legislative history, which makes it clear that local hazards can not be “statewide in character” and that “there is no intent to permit a State to establish Statewide standards superimposed on national standards covering the same subject matter.” H.R.Rep. 91-1194 (1970), reprinted in 1970 U.S.C.C.A.N. 4104, 4117. The Court rejects, however, plaintiffs’ and DOT’s suggestion that the local safety hazard must be “very nearly, if not completely, unique,” see DOT’s brief at 18 — at least to the extent that the term “unique” is defined to require a one-of-kind, or virtually one-of-a-kind, hazard. First, the statutory language does not employ the term “unique.” Instead, it simply states that the regulation must be necessary to reduce an “essentially” local safety hazard, a more generous term. Notably, while Congress added the term “essentially” to qualify the phrase “local safety hazard,” it declined the railroad’s invitation to adopt the term “unique.” See Railroad Safety and Hazardous Materials Control Act of 1970: Hearings on H.R.7068, H.R.14417 and H.R.14478; Before the Subcomm. On Transp. and Aeronautics of the Comm, on Interstate & Foreign Commerce, 91st Cong. 2nd Sess., Ser. No. 91-51, p. 100 (April 22, 1970) (proposing that term unique be used to qualify “local safety hazard”).” An insistence on true “uniqueness” would effectively eviscerate the exception, a result which is not supported by the legislative history. See H.R.Rep. 91-1194, reprinted in 1970 U.S.C.C.A.N. at 4117 (“The States will retain authority [pursuant to the local safety hazard provision] to regulate individual local problems where necessary to eliminate or reduce essentially local railroad safety hazards.”). Second, the railroads’ argument, that local safety hazards must be truly unique in light of other language in the savings clause, is not convincing. As plaintiffs’ vigorously emphasize, the savings clause provides in part that the laws “related to railroad safety shall be nationally uniform to the extent practicable.” 49 U.S.C. § 20106. While national uniformity is clearly an important objective, the paramount purpose of the Act is railroad safety. 49 U.S.C. § 20101 (“The purpose of this chapter is to promote safety in every area of railroad operations ... ”); see also H.R.Rep. No. 91-1194, reprinted in 1970 U.S.C.C.A.N. at 4104 (“The primary purpose of this legislation ... is to promote safety ...”); CSX, 507 U.S. at 661, 113 S.Ct. 1732. Hence, Congress’ qualification that the laws relating to railway safety shall be nationally uniform “to the extent practicable. ” In short, the FRSA expresses a desire to achieve both safety, and, to the extent practicable, uniformity — not an intent to trump safety with uniformity. Nor are we persuaded that Congress intended, as plaintiffs suggest, to limit the exception to “industrial or plant railroads.” Had this been the intention, it could have stated so quite simply. Third, the Court agrees with DOT that a site is not a “local” safety hazard simply because an accident occurs there; rather there must be a locally rooted cause. Indeed, every accident occurs somewhere “local,” but that does not make it a local safety hazard. The Court disagrees, however, that the locally rooted cause must be limited to some extremely unusual or unique geological anomaly that itself primarily causes an accident. While the hazard must be limited to a localized or geographically discrete area, Congress nowhere required that the hazard itself must be viewed solely in terms of unique geology or geography. Certainly, if an accident happens at a particular location, but for reasons completely unrelated to the location, the accident can not establish the location as a local safety hazard. It is also true, however, that particular and uncommon features of a specific location can substantially increase the risk, and the severity of the risk, even if there is no “unique” geological anomaly present. For example, if a specific stretch of track features an unusually severe grade and curve combination close to an unusually important waterway, or exceptionally densely populated area, it may be much more likely that (a) a bad brake will fail at that particular point, and (b) that such an accident will have unusually catastrophic consequences. The fact that it is the “external” bad brake that is a primary cause of the accident, as opposed to some unique geological feature, does not detract from the fact that it is the local conditions that combine — in any common sense meaning of the term- — to create an “essentially local safety hazard.” Accordingly, this Court concludes that the phrase “essentially local safety hazard” is intended to strike a balance — it encompasses more than a truly unique condition or geological anomaly but plainly less than authority to superimpose statewide regulations. Rather, the phrase is directed at geographically discrete, localized areas that have peculiar or distinctive features or characteristics (or combinations thereof), that are neither typical nor common or otherwise “state-wide” in nature, and which create a safety hazard. Such an approach appears most consistent with both the language and spirit of the FRSA, as well as its legislative history. See e.g. Monongahela, 404 A.2d at 1379-80 (finding that state rule imposing special precautions for single blind curve, is “purely local and is designed to reduce or eliminate a purely local safety hazard”). (Q) Site 9 Shasta Line (Black Butte District), Milepost 322.1 to 332.6 Site 9, known as the Cantara Loop, is a 10.5 mile stretch of track in the mountains in Northern California starting at a point north of Dunsmuir and ending at a point below Mt. Shasta. This line goes through the Sacramento River Canyon and crosses the river at several locations. This site experienced 32 out of 90 derailments on the line between 1976 and 1991, although there have been no accidents since 1992. Decision at 127; Corrected Site Exhibit Book (“SEB”) at site tab 9. “Each statistical method confirms that the chance of the accident concentration at this site occurring randomly is extremely small at less than 1 in a trillion.” Decision at 127 (footnote omitted). The particularly sharp 14 percent curve is the sharpest main line track curve in the state of California. Id. at 128. The curve and grade have been directly linked to one-third of the derailments at the site. Id. at 127. Track-train dynamics forces also contributed to many other derailments. Id. at 128. The most hazardous grade-curve combination occurs on a bridge which crosses over the Sacramento River, a major supplier of the state’s water. This resource-critical river is already damaged from the devastating environmental effects of the 1991 Dunsmuir accident which poisoned the river and destroyed its ecosystem for many miles. King Decl. at ¶ 137. The severity of the environmental risk from future accidents can only be described as enormous. Of course, any such risk also necessarily entails serious risks to the physical and economic health of the surrounding communities as well. While the grade and curve at site 9 may not be unique in the sense that there are a few other curves in other states that are equal to or slightly higher (there is a 16 percent curve in Colorado and 14 percent curves in Idaho), California need not demonstrate that the grade and curve of the site is uniquely different from every other site in the country. Rather, it must show that the site is a geographically discrete, localized area, that exhibits a combination of peculiar or distinctive features or characteristics (including environmental or demographic features) that are neither typical nor common or otherwise “state-wide” in nature, and which create a safety hazard. Given all of the above, the Court is satisfied that site 9 satisfies this standard and thus constitutes an “essentially local safety hazard.” (Ü) Necessity The Court is also satisfied that the rule is necessary to reduce or minimize the local safety hazard. Plaintiffs themselves have concluded that strengthening the track is needed to reduce the risk and prevent a return to historically high derailment rates. As an SP representative previously testified during hearings following the 1991 Dunsmuir accident: As a result of [track strengthening] measures, we developed a much stronger track structure [at site 9], better able to resist sudden lateral transient forces.... The measures that we took to strengthen the track structure, and to help the train engineers in their descending grade train handling, were successful in reducing, and virtually eliminating, the cluster of derailment problems suggested from the FRA statistics. Decision at 90. See also id. at 89, 128. Plaintiffs argue that the rule is unnecessary because UP, which currently operates the trains over site 9, has voluntarily exceeded the minimum federal standards and there is no evidence that it is likely to abandon this effort. There is no guarantee, however, that the necessary standards will always be voluntarily maintained either by UP or future operators. Furthermore, the critical inquiry is not whether the current operator asserts an intent to voluntarily comply, but whether the higher track strength is necessary to reduce the local safety hazard. (Hi) Incompatibility with federal law California’s track standard rule for site 9 is not incompatible with the federal track standards simply because it is more stringent. On the contrary, the FRSA expressly states that the state may adopt a “more stringent law” where necessary to reduce a local safety hazard. 49 U.S.C. § 20106. Plaintiffs, however, argue that the state rule is incompatible with federal track standard rules because the state rule prohibits UP from altering the standards without first obtaining approval. This provision, plaintiffs argue, conflicts with 49 C.F.R. § 213.1(a), which provides in part that federal regulations governing minimum track safety standards are not intended “to restrict a railroad from adopting and enforcing additional or more stringent requirements.... ” The suggestion that California’s rule will somehow prevent plaintiffs from implementing stricter standards is no more than a manufactured conflict. The purpose of the California rule is to promote stricter standards, not prevent them. See Decision at 91 (“We assure the Railroads that our regulations will never preclude a Railroad from implementing tougher track standards than the minimum standards we intend to require”). Indeed, the existing track strength exceeds the requirements set forth by CPUC’s rule; yet, California has always “accepted this track as being in compliance with the Decision’s requirements.” Pendleton Decl. ¶ 11. Plaintiffs also argue that they will be discouraged from making voluntary improvements if such actions could readily lead to mandatory rules. This argument is equally meritless. First, a voluntary effort can only be made mandatory in the event that a state is able to demonstrate that the rule is necessary to reduce or eliminate a local safety hazard. In such a circumstance, the railroad can hardly complain that such a rule should not be mandatory. Second, plaintiffs have made no showing that their hypothetical concern rises to the level of a conflict with a federal rule. Accordingly, this Court finds that the track standard rule is not incompatible with federal law. (iv) Unreasonable burden on interstate commerce Plaintiffs do not contend that the track standards rule will unreasonably burden interstate commerce; nor does the record reflect any such burden. Accordingly, defendants’ track standard rule is saved from preemption pursuant to the local safety hazard provision of 49 U.S.C. § 20106. 3. Conclusions regarding preemption under the FRSA In light of the above, the Court finds that the following CPUC rules are preempted by the FRSA because they are covered by federal regulations and do not satisfy the requirements of the local safety hazard savings clause: (1) Rule requiring CPUC approval of changes to railroad operating train make-up rules (2) Rule requiring additional training on train make-up rules The Court finds that the following CPUC rules are not preempted by the FRSA because they are not covered by federal regulations: (1) Rule requiring railroads to cooperate in developing and implementing train make-up rules based on current TTD principles, (2) Rule requiring railroads to comply with their own operating rules concerning train make-up (3) Rule requiring trackside defect detectors (Hot box) Rule The Court finds that the following CPUC rule is not preempted by the FRSA, although it is covered, because it satisfies the requirements of the local safety hazard exception: (1) Track standards rule B. THE LOCOMOTIVE BOILER INSPECTION ACT The Locomotive Boiler Inspection Act (“LBIA”), 49 U.S.C. § 20701 et seq., prohibits carriers from using a locomotive unless “the locomotive or tender and its parts and appurtenances ... are in proper condition and safe to operate without unnecessary danger of personal injury ...” 49 U.S.C. § 20701. Although the LBIA is silent as to its preemptive effect, it has long been held to constitute a total occupation of “the field,” preempting any state or local regulation on the same subject. See Napier v. Atlantic Coast Line R.R., 272 U.S. 605, 612-13, 47 S.Ct. 207, 71 L.Ed. 432 (1926). The scope of “the field” is governed by the general power delegated by the BIA which extends to “the design, the construction and the material of every part of the locomotive and tender and of all appurtenances,” id. at 611, 47 S.Ct. 207; Southern Pac. Transp. Co., 9 F.3d at 811, as well as maintenance of locomotives and their parts. 49 U.S.C. § 20701. The purpose of the broad preemptive sweep of the BIA is to maintain nationally uniform equipment standards, so that “locomotive companies need only concern themselves with one set of equipment regulations and need not be prepared to remove or add equipment as they travel from state to state.” Law v. General Motors Corp., 114 F.3d 908, 910 (9th Cir.1997) (quoting Southern Pac. Transp. Co., 9 F.3d at 811). The Court addresses in turn the CPUC rules challenged under the LBIA. 1. TTD Train Make-up Rules The TTD train make-up rules not preempted by the FRSA concern (1) the development of performance based train make-up rules for 13 sites, and (2) compliance with the railroads’ own make-up rules for those sites. Although train make-up only concerns the configuration of the train, and not any train equipment, plaintiffs nonetheless contend that all of the train make-up rules are preempted by the LBIA. This argument must fail because the TTD rules at issue do not, as the railroads suggest, force them to use high-strength couplers or otherwise alter the design, construction or material of the trains or their appurtenances;