|SURFACE TRANSPORTATION BOARD DECISION DOCUMENT|
|CARGILL, INCORPORATED V. BNSF RAILWAY COMPANY|
|DECISION: (1) DENIED CARGILL’S REQUEST TO SUPPLEMENT ITS COMPLAINT WITHOUT PREJUDICE; AND (2) DISMISSED THE COMPLAINT AND DISCONTINUED THIS PROCEEDING.|
| 199 KB|
|Approximate download time at 28.8 kb: 1 Minutes|
If you do not have Acrobat Reader, or if you have problems reading our files with your current version of Acrobat Reader, the latest version of Acrobat Reader is available free at www.adobe.com.
|Full Text of Decision|
42719 SERVICE DATE – AUGUST 12, 2013
SURFACE TRANSPORTATION BOARD
Docket No. NOR 42120
CARGILL, INCORPORATED v. BNSF RAILWAY COMPANY
Digest: BNSF Railway Company (BNSF) imposes a fuel surcharge on the tariff rates it charges its customers for carload shipments of agricultural and industrial products. Cargill, Inc., one of BNSF’s customers, filed a complaint against BNSF, alleging that its fuel surcharge has violated the Board’s rules and constitutes an unreasonable practice. This decision finds that Cargill did not show that BNSF’s fuel surcharge constitutes an unreasonable practice under current fuel surcharge rules. However, the Board also gives notice that it will start a proceeding to obtain public comments on the Board’s “safe harbor” rule, which allows rail carriers to rely on a Board-approved fuel index to measure changes in fuel prices for purposes of their fuel surcharge programs.
Decided: August 9, 2013
Cargill, Incorporated (Cargill) has filed a complaint challenging fuel surcharges collected by BNSF Railway Company (BNSF) between April 19, 2008 and December 31, 2010, as an unreasonable practice under 49 U.S.C. § 10702(2). Cargill requests that the Board: (1) find the challenged fuel surcharge practices to be unreasonable and order BNSF to cease and desist from such practices; (2) prescribe reasonable fuel surcharge practices; and (3) award monetary damages with interest under 49 U.S.C. § 11704(b) for all unlawful fuel surcharge payments made. In this decision, we are finding that Cargill has failed to establish that the challenged fuel surcharge practices are unreasonable. As a consequence, we are dismissing Cargill’s complaint and discontinuing the proceeding.
We reach this result in large measure from the operation of the “safe harbor” provision of the Board’s current fuel surcharge rules, adopted in Rail Fuel Surcharges (Fuel Surcharges), EP 661, slip op. at 11 (STB served Jan. 26, 2007), under which rail carriers may measure changes in their fuel costs by a particular fuel index endorsed by the Board. Because we are concerned that there may be outcomes that the Board did not anticipate when the safe harbor provision was adopted, we are also giving notice that the Board will be issuing an advanced notice of proposed rulemaking to give shippers, rail carriers, and other interested parties the opportunity to comment on the safe harbor provision, including whether it should be modified or removed.
BNSF’s Adoption of a Mileage-Based Fuel Surcharge. Around 2001, when diesel fuel prices began to soar, many rail carriers, including BNSF, adopted fuel surcharge programs. For ease of administration, rail carriers calculated the fuel surcharges they assessed on shipments as a percentage of the base rate. In early 2005, in response to shipper complaints, BNSF replaced certain of its existing percent-of-rate fuel surcharges with mileage-based fuel surcharges. BNSF intended to implement a mileage-based fuel surcharge for agricultural and industrial commodities at the start of 2006, but customers in its Industrial Products group needed more time to adjust their information systems to handle the proposed mileage-based fuel surcharge program. As a result, BNSF implemented its new mileage-based fuel surcharge only on agricultural products beginning in January 2006, and extended it to all industrial products in April 2007. BNSF also implemented a separate, mileage-based fuel surcharge on unit train coal and taconite traffic beginning in January 2006.
The Board’s Fuel Surcharge Rules. While BNSF was transitioning to mileage-based fuel surcharges, the Board also began to receive complaints from shippers about percent-of-rate fuel surcharges still common among other railroads. On May 11, 2006, the Board held a hearing to allow rail shippers, rail carriers, and other interested parties to comment on how rail carriers calculate and charge fuel surcharges. Following the hearing, the Board proposed to: (1) require that rail fuel surcharges “be tied not to the level of the base rate but to those attributes of a movement that directly affect the amount of fuel consumed,” such as mileage or mileage and weight; (2) prohibit “double-dipping,” which it defined as charging a fuel surcharge while also increasing the base rate with a rate escalator (such as the Board’s Rail Cost Adjustment Factor) that includes a fuel-cost component; (3) require rail carriers to use a particular fuel index—the Energy Information Administration’s (EIA) “U.S. No. 2 Diesel Retail Sales by All Sellers (Cents per Gallon),” now referred to as the Highway Diesel Fuel Index (HDF Index)—when calculating changes in their fuel costs for purposes of their fuel surcharge programs; and (4) require large rail carriers to file with the Board periodic reports on their total fuel costs, gallons consumed, and fuel surcharge revenues.
After receiving dozens of comments in Fuel Surcharges, the Board issued its final rules on January 26, 2007. First, the Board prohibited rate-based fuel surcharges. The Board explained that, consistent with the rail transportation policy “to encourage honest and efficient management of railroads,” 49 U.S.C. § 10101(9), its new rules were “only addressing the manner in which railroads apply what they label a fuel surcharge.” Fuel Surcharges, slip op. at 7. The Board emphasized that it was not limiting the total rate rail carriers could charge through a combination of base rates and surcharges for providing rail transportation. Rather, the Board stated, that it was only addressing the manner in which rail carriers apply what they label a fuel surcharge. Id. Rail carriers could still raise their rates if they wished to do so, subject to the statutory rate reasonableness standard, but could not impose rate increases on the basis of a misrepresentation. Second, noting that nearly every commenter supported the proposed prohibition of “double-dipping,” the Board banned that practice. Id. at 10. Third, explaining that it could not conclude that the use of all other fuel indices was unreasonable, the Board decided not to mandate the use of the HDF Index. Instead, the Board stated that the HDF Index “is a reasonable index to apply to measure changes in fuel costs for purposes of a fuel surcharge program. Thus, it provides a ‘safe harbor’ upon which carriers can rely for an index.” Id. at 11. Finally, the Board decided to go forward with its reporting proposal, but to do so in a separate rulemaking proceeding to allow for comments on the costs and burdens of reporting to comport with the Paperwork Reduction Act, 44 U.S.C. § 3501 et seq. In August 2007, in the separate proceeding, the Board amended its regulations at 49 C.F.R. § 1243.3 to require Class I rail carriers to report certain data concerning fuel costs and fuel surcharges billed.
In Dairyland Power Cooperative v. Union Pacific Railroad (Dairyland), NOR 42105 (STB served July 29, 2008), the first complaint case to challenge a specific rail fuel surcharge program under the Board’s fuel surcharge rules, Dairyland Power Cooperative (Dairyland) alleged that the fuel surcharges collected by the defendant rail carrier under its mileage-based fuel surcharge program exceeded its increased fuel costs and were generating substantial profits. The Board, in denying a motion to dismiss, pointed out that if the rail carrier was deriving substantial profits from fuel surcharges “that could in turn call into question the reasonableness of [the rail carrier’s] fuel surcharge program.” Id., slip op. at 5. The Board cautioned, however, that it “cannot expect a precise match between fuel surcharge revenues and increased fuel costs for any one shipper.” Id. Rather than requiring a fuel surcharge program to take into account “every conceivable factor that could affect fuel costs,” the Board stated that it requires “only that any fuel surcharge program a railroad uses ‘must be based on the attributes of a movement that directly affect the amount of fuel consumed.’” Id. The Board clarified that to establish an unreasonable practice a complainant must show that “the general formula used to calculate fuel surcharges bears no reasonable nexus to the fuel consumption for the traffic to which the surcharge is applied.” Id. at 6.
This Case. On April 19, 2010, Cargill, a major shipper of agricultural products, filed a complaint challenging aspects of BNSF’s fuel surcharge program. Cargill ships various agricultural and other commodities over BNSF in common carrier service under a number of BNSF pricing authorities. In addition to the assessed linehaul rate, BNSF charges, and Cargill pays, a mileage-based fuel surcharge on all carload shipments of agricultural and industrial products under BNSF Rules Book 6100-A, Item 3375L, Section B. Section B is incorporated by reference into BNSF's common carrier pricing authorities applicable to Cargill's traffic and became effective with respect to agricultural and industrial products in January 2006 and April 2007, respectively. The fuel surcharge is “calculated by multiplying the applicable fuel surcharge per mile times the number of miles per shipment.” Complaint, Exhibit A at 41, BNSF Rules Book 6100-A, Item 3375L, Section B. The amount of the fuel surcharge per mile can change monthly. It is based on the HDF Index, which, as noted, is published by the EIA, an independent arm of the U.S. Department of Energy, and is available on BNSF’s website in the form of a weekly average price. The surcharge per mile to be applied in any given month is based on the average monthly HDF price published two months prior to the month of assessment. According to BNSF, this is the minimum time lag necessary to permit publication. Average HDF prices for the immediately preceding month cannot be used because they are not published in time.
The challenged fuel surcharge appears in BNSF’s rules tariff as a table that sets out in two columns the monthly average HDF prices in 4¢ increments and the corresponding fuel surcharge in cents per mile. The surcharge starts at 1¢ per mile and increases by 1¢ per mile for every 4¢ per gallon increase in the monthly average HDF price (the so-called “1:4 step function”). This is because the surcharge is based on the assumption that a loaded car on average moves at 4 MPG. The starting point for assessing the surcharge, referred to as the “strike price,” is the HDF price of $1.25 per gallon. Thus, an average HDF price at the time of shipment of between $1.25 and $1.289 per gallon results in a fuel surcharge of 1¢ per mile; an average HDF price of between $1.29 and $ 1.329 per gallon results in a surcharge of 2¢ per mile, and so on in 4¢ per gallon increments. To calculate the surcharge on a shipment, a shipper may go to BNSF’s website to obtain the published average HDF price for the shipment date and then consult BNSF’s tariff for the corresponding fuel surcharge in cents per mile. That surcharge in cents per mile is then multiplied by the shipment’s mileage to obtain the actual surcharge on a shipment. Mileage is the shortest rail mileage between origin and destination, which is posted on BNSF’s website in its Mileage Inquiry Tool.
Cargill claims that BNSF’s collection of fuel surcharges constitutes an unreasonable practice under § 10702(2), asserting in its Complaint at 3-4 that:
· “the general formula [used] to calculate the fuel surcharges bears no reasonable nexus to, and overstates, the fuel consumption for the BNSF system traffic to which the surcharge is applied” (the “Reasonable Nexus” claim);
· “BNSF is using [the fuel surcharge] to extract substantial profits over and above its incremental fuel costs for the BNSF system traffic to which the surcharge is applied” (the “Profit Center” claim); and
· “BNSF is double recovering the same incremental fuel cost increases BNSF has incurred in providing common carrier service to Cargill” by capturing the same fuel cost increases in both the base rate and the fuel surcharge (the “Double Recovery” claim).
Cargill seeks damages plus interest for shipments that moved between April 19, 2008 (two years before it filed its complaint) and December 31, 2010 (the end date of Cargill’s cost analysis). If an unreasonable practice is found, Cargill asserts that the Board’s decision should apply to shipments that moved, and continue to move, after December 31, 2010, under the challenged fuel surcharge tariff and subsequent iterations. See supra note5.
BNSF filed an answer to the complaint on May 10, 2010, and a motion for partial dismissal on May 28, 2010. In that motion, BNSF argued that under the principles established in Fuel Surcharges, Dairyland, and Union Pacific Railroad v. ICC (Union Pacific), 867 F.2d 646 (D.C. Cir. 1989), the Board should dismiss Cargill’s Profit Center and Double Recovery claims and its request for damages with interest on all three claims. Cargill filed a reply on June 17, 2010.
In a decision served on January 4, 2011, the Board denied BNSF’s motion to dismiss with respect to the Profit Center claim and granted it with respect to the Double Recovery claim. The Board also found it premature to rule on BNSF’s argument that Cargill cannot recover damages under any of its claims because any alleged injury would amount to a claim that the fuel surcharge resulted in unreasonable rates, contrary to Union Pacific, and it adopted a procedural schedule. Cargill, on January 24, 2011, filed a petition for reconsideration of the 2011 Decision to the extent it dismissed Cargill’s Double Recovery claim. In a decision served on May 25, 2012, the Board denied this petition.
Meanwhile, the parties filed opening, reply, and rebuttal statements. Thereafter the Board, in a decision served on March 1, 2012, granted BNSF’s request for the simultaneous filing of final briefs and directed the parties to address two additional questions concerning the standard for determining whether a fuel surcharge program constitutes an unreasonable practice and at what point in time a fuel surcharge program becomes an unreasonable practice. The parties filed their final briefs on April 2, 2012.
DISCUSSION AND CONCLUSIONS
This case, the first challenge to a specific rail fuel surcharge to be decided by this agency, raises a host of novel issues. As explained in more detail below, we conclude that Cargill has failed to show that the mileage-based fuel surcharge that BNSF applied to all shipments of agricultural and industrial products moving under tariff from 2006 through 2010 constituted an unreasonable practice under current fuel surcharge rules. However, we recognize that the safe harbor HDF Index may lead to outcomes that the Board did not anticipate when that provision was adopted in Fuel Surcharges.
In Part I, we explain that Cargill has failed to show that BNSF over-recovered its incremental fuel costs as properly measured by the safe harbor HDF Index. This finding, however, does not end the inquiry. Cargill also claims that BNSF designed the fuel surcharge knowing that it would lack a reasonable nexus to fuel consumption. If that were so, the lack of any over-recovery could simply be happenstance, and BNSF could still be found to have engaged in an unreasonable practice and be directed to adopt a reasonable surcharge mechanism. In Part II, we address the reasonable nexus issue and explain that Cargill has failed to show that the challenged fuel surcharge bears no reasonable nexus to BNSF’s fuel consumption. Finally, in Part III, we explain that in a future proceeding the Board will seek public input on the safe harbor provision, including whether it should be modified or removed.
I. The Profit Center Claim
Cargill contends that BNSF’s collection of fuel surcharges constitutes an unreasonable practice under § 10702(2), asserting that BNSF has been using the fuel surcharge to extract substantial profits over and above its incremental fuel costs for the BNSF system traffic to which the surcharge has been applied. Using BNSF’s internal fuel costs, not HDF Index prices, Cargill developed an incremental fuel surcharge study that made movement-specific cost calculations for the more than 5.6 million individual shipments subject to the challenged fuel surcharge that BNSF moved for Cargill and BNSF’s other shippers between January 1, 2006 and December 31, 2010 (the five-year period) and compared the resulting fuel costs to the over $2 billion in fuel surcharge revenues that BNSF collected on those shipments. Based on that study, Cargill asserts that BNSF over-recovered its incremental fuel costs by $560.9 million. BNSF disagrees, arguing that it under-recovered its incremental fuel costs by almost $14 million as measured by the HDF Index. While the parties do not dispute the amount of revenue that BNSF collected under the challenged fuel surcharge program, a central disagreement in the case is how to measure BNSF’s incremental fuel costs.
The Board will use the HDF Index to measure BNSF’s incremental fuel costs in this case. As we explain in more detail below, in Fuel Surcharges the Board adopted the HDF Index as a safe harbor. That is, the Board allowed rail carriers to use the HDF Index to measure their incremental fuel costs. The Board determined that: (1) the HDF Index would provide a simple and transparent methodology for measuring incremental fuel costs, one that would not be subject to manipulation; and (2) the markets for fuel are all highly correlated, so, over time, changes in the HDF Index would be likely to match quite closely changes in the actual price of the fuel that rail carriers purchase.
Because the Board previously announced that the HDF Index would be a safe harbor to measure incremental fuel costs, BNSF was justified in relying on it to do so. BNSF adopted the HDF Index for the challenged fuel surcharge and set its strike price at $1.25 per gallon. Therefore, the proper way to measure the change in BNSF’s monthly fuel costs is simply to subtract the $1.25 strike price from the monthly HDF price per gallon.
If changes in the price of fuel purchased by BNSF are measured based on the difference between the HDF price and the $1.25 strike price, then to determine whether BNSF over- or under-recovered its incremental fuel costs, one would multiply the incremental fuel costs by the gallons consumed. Then the resulting incremental fuel costs would be compared to BNSF’s fuel surcharge revenues. As we discuss below, we examined the BNSF fuel cost study that used the HDF Index to measure the monthly changes in its incremental fuel costs, and concurring that BNSF has not over-recovered, we conclude on that basis that Cargill’s Profit Center claim must fail. We also resolve other issues disputed by the parties in connection with the Profit Center claim, including whether: (1) movement-specific adjustments to the Uniform Rail Costing System (URCS) may be made in calculating BNSF’s fuel costs; (2) the effects of hedging should be considered in calculating BNSF’s fuel costs; and (3) BNSF may use its fuel surcharge program to recover incremental locomotive unattributable and non-locomotive fuel costs.
The Safe Harbor. In Fuel Surcharges—Proposed, the Board noted that “shippers argue nearly unanimously that transparency is needed for rail fuel surcharges,” and on that basis the Board proposed that rail carriers be required to use the HDF Index to measure changes in their fuel costs. Fuel Surcharges—Proposed, slip op at 2. Following notice and comment, the Board adopted the HDF Index (then referred to as the EIA Index) as a safe harbor, stating as follows:
Strong support has been expressed in the record for the proposal that railroads apply a single, uniform index to measure changes in fuel prices. Shippers argue that it would better ensure accuracy, transparency and accountability, and thereby enhance the credibility of fuel surcharges in the eyes of those who pay them. Moreover, there is general agreement─even among those carriers that object to Board imposition of a uniform index─that the EIA Index accurately reflects changes in fuel costs in the rail industry. Indeed, the EIA Index closely correlates with other fuel cost indices, including the indices currently used by most carriers. Moreover, the [Association of American Railroads] has developed an index for carriers that is virtually identical to the EIA Index.
Because the EIA Index has been the subject of notice and comment and has withstood scrutiny on this record . . . we conclude that it is a reasonable index to apply to measure changes in fuel costs for purposes of a fuel surcharge program. Thus, it provides a “safe harbor” upon which carriers can rely for an index. Use of an alternative index may be subject to challenge.
. . . [A]ny alternative index used may be challenged as unreasonable on a case-by-case basis, and should meet or exceed the standards of accuracy, transparency, availability, and neutrality of the EIA Index, should closely correlate with other indices, and should reflect fuel price changes quickly.
Fuel Surcharges, slip op. at 11 (emphasis added).
The Board intended to encourage use of the HDF Index by assuring rail carriers that they could rely on changes in the HDF Index as a proxy for measuring changes in their internal fuel costs. The safe harbor does not, however, immunize a rail carrier’s fuel surcharge program from challenge simply because the program uses the HDF Index. A fuel surcharge program employing the HDF Index could still be shown to be unreasonable if other aspects of the program are unreasonable. For example, if a mileage-based fuel surcharge program rests on an unreasonable assumption about the fuel efficiency of the traffic to which it applies, then the Board could find that the program lacks a reasonable nexus to fuel consumption. Indeed, Cargill makes just such a claim here, as well as several others challenging the reasonableness of various aspects of BNSF’s fuel surcharge program, all of which we address later in this decision. However, what the safe harbor means is that if a rail carrier uses the HDF Index to measure changes in its fuel costs, then that is how the Board will measure these changes as well, rather than by looking at evidence of changes in the rail carrier’s internal fuel costs.
Cargill asserts that the use of the HDF Index to measure incremental fuel costs is contrary to the Board’s 2011 Decision, which denied BNSF’s motion to dismiss Cargill’s Profit Center claim and granted BNSF’s motion to dismiss Cargill’s Double Recovery claim. That decision, according to Cargill, “violates governing legal standards [which] call for the parties to measure the difference between BNSF’s fuel surcharge revenues and its ‘actual incremental cost of fuel.’” Cargill Rebuttal Statement at 46 (citation omitted). Cargill argues that “BNSF was not paying incremental HDF prices for fuel, it was paying lower actual incremental fuel prices [and as a consequence the inclusion of fuel prices BNSF was not paying has no place in calculating BNSF’s ‘actual incremental cost of fuel.’” Id. at 46.
Cargill interprets the Board’s safe harbor as merely providing “that if a shipper filed a fuel surcharge complaint . . . and the carrier was using an HDF index, the shipper could not ask the Board to prescribe the use of a different index.” Id. at 47. Cargill claims that the Board endorsed the use of the HDF Index as a reasonable way “to estimate relative change in railroad fuel costs but not as a substitute for absolute change in railroad fuel cost.” Cargill Rebuttal Statement, Crowley/Mulholland V.S. at 47 (citation omitted). Based on a regression analysis of data from 2006 to 2010, Cargill maintains that if BNSF uses the HDF Index in its fuel surcharge, it must use a “consumption adjustment factor” in developing the step function, Cargill Opening Statement, Crowley/Mulholland V.S. at 25; Cargill Rebuttal Statement, Crowley/Mulholland V.S. at 50, to reflect the fact that “for every one cent change in HDF price, there is a 0.924 cent change in BNSF fuel price.” Cargill Opening Statement, Crowley/Mulholland V.S. at 25.
Cargill’s interpretation of the safe harbor is flawed in two basic respects. First, it is contrary to the plain wording of the Board’s Fuel Surcharges decision. There, the Board said nothing about relative changes in fuel prices. Instead, the Board said that the HDF Index “is a reasonable index to apply to measure changes in fuel costs for purposes of a fuel surcharge program.” Fuel Surcharges at 11. Cargill improperly relies on the Board’s 2011 Decision to argue that, despite the safe harbor discussion in Fuel Surcharges, BNSF’s internal fuel costs must be used to analyze Cargill’s Profit Center claim. The language Cargill cites from the Board’s 2011 Decision, see supra note 16, did not withdraw retroactively and without notice the safe harbor, which had been adopted after a lengthy notice and comment process in the Fuel Surcharges proceeding. The Board’s 2011 Decision did not discuss the safe harbor HDF Index at all, and the issue was not raised by the parties in either the motion to dismiss or the reply. Moreover, when discussing design flaws that could lead to a finding that BNSF’s fuel surcharge program was unreasonable, the Board did not reference the HDF Index. Second, if accepted, Cargill’s interpretation of the safe harbor provision would rob the safe harbor of any real meaning. Practically speaking, rail carriers would not be able to rely on the HDF Index to estimate “relative” changes in fuel costs because, to come up with the correct adjustment factor, they would have to be able to predict precisely how changes in that index would correlate with future changes in their own internal fuel costs. Although the correlation would likely be high, it would never be perfect and could not be predicted with certainty. Hence, the Board uses the HDF Index to measure BNSF’s incremental fuel costs.
Movement-Specific Adjustments to URCS. Cargill’s incremental fuel cost study sought to calculate the rail carrier’s variable locomotive fuel costs as measured by URCS. In the process, Cargill made piecemeal movement-specific adjustments to URCS, allegedly to more accurately reflect BNSF’s variable locomotive fuel costs. Cargill’s movement-specific adjustments to URCS are part of its effort to measure BNSF’s incremental fuel costs by employing BNSF’s internal fuel costs. We reject Cargill’s attempt to rely on movement-specific adjustments in this case. First, as discussed above, the appropriate way to measure a rail carrier’s incremental fuel costs is to rely on the difference between the safe harbor HDF Index, which serves as a proxy to measure the rail carrier’s incremental fuel costs, and the $1.25 strike-price. Second, the Board no longer permits movement-specific adjustments to URCS in rate reasonableness cases, and it has recently declined to allow movement-specific adjustments in a non-rate context. As in those contexts, we conclude that permitting piecemeal movement-specific adjustments to URCS in the fuel surcharge context, as Cargill proposes, would not likely lead to more accurate results, and would almost certainly increase litigation and litigation costs.
Fuel Hedging. In its incremental fuel cost study, Cargill also adjusts BNSF’s fuel costs by its fuel hedging gains. Like movement-specific adjustments to URCS, however, fuel hedging gains are relevant only to measure the incremental fuel costs based on internal fuel costs incurred by a rail carrier. Accordingly, we need not consider fuel hedging gains given that the safe harbor HDF Index is the appropriate way to measure BNSF’s incremental fuel costs under the Board’s current fuel surcharge rules. Furthermore, to account for hedging gains and losses, rail carriers would have to figure out a way to adjust retroactively upwards and downwards previously paid surcharge bills, a costly and extremely complex task. In addition, hedging gains could not be considered without considering hedging losses, and, in that respect, we do not believe it would be fair for shippers to pay for a rail carrier’s hedging losses.
Finally, Cargill’s attempt to measure BNSF’s incremental fuel costs by reference to changes in BNSF’s internal fuel costs and offsetting them with BNSF’s fuel hedging gains is also inconsistent with the principles of uniformity, transparency, public availability, and neutrality advocated by shippers in the Fuel Surcharges proceeding. Without an accurate, publicly available fuel surcharge index, there would be no way, other than by filing a rate complaint case under § 10702(2), to ensure that a rail fuel surcharge accurately reflects the rail carrier’s increased cost of fuel.
Locomotive Unattributable and Non-Locomotive Fuel Costs. Cargill’s incremental fuel cost study also excluded BNSF’s incremental locomotive unattributable and non-locomotive fuel costs. We disagree with Cargill’s contention that these fuel costs must be excluded in calculating monthly gallons consumed. These fuel costs are legitimate and necessary expenses, and they are affected just like variable locomotive fuel costs when the price of fuel rises. Rail carriers are permitted to recover these increased expenses. Cargill’s suggestion that these increased fuel costs must be recovered in the base rates would defeat the purpose of having a fuel surcharge program in the first place. The reason for a fuel surcharge program is to make it possible for rail carriers to recover their incremental fuel costs in a volatile market without having to revise their base rates to account for every spike or dip in the price of fuel. Yet, if we were to follow Cargill’s approach, rail carriers would have to do just that to fully recover (but not over-recover) their incremental fuel costs.
According to Cargill, the law requires this implausible result. Cargill argues that, under Board precedent, rail carriers may impose a fuel surcharge to recover only “incremental fuel costs attributable to the movement involved.” Cargill Rebuttal Statement at 23 (citing Fuel Surcharges—Proposed at 5; Dairyland at 1; and 2011 Decision at 5). Thus, Cargill reasons, because locomotive unattributable and non-locomotive fuel costs cannot be attributed to any particular movement, a rail carrier cannot recover those costs through a fuel surcharge.
Cargill reads our cases too broadly. Since becoming aware of the shipping community’s dissatisfaction with the railroad industry’s method of assessing fuel surcharges, the Board has sought to prohibit the highly unfair outcomes that are possible under rate-based fuel surcharges. As we explained:
[T]ying the level of a fuel surcharge to the level of the base rate means that even two shippers that have shipments moving over the same route between the same points could be charged vastly different fuel surcharges, notwithstanding the fact that the fuel costs associated with their movements could be identical. And it means that a short-haul captive shipper would pay a much higher fuel surcharge than a longer-haul competitively served shipper of the same commodity even though the total fuel costs for the captive shipper’s movement would be less because of the differential pricing of the base rates.
Fuel Surcharges—Proposed at 4; see also Fuel Surcharges at 6 (“Two shippers may have traffic with identical fuel costs, but if one starts out with a higher base rate (because, for example, it has fewer transportation alternatives), it will pay dramatically more in fuel surcharges.”). The Board sought to prevent rail carriers from forcing one set of shippers (mainly captive shippers) to pay the higher fuel bills that could more rationally be attributable to another set of shippers (competitively served shippers). In contrast, the incremental fuel costs at issue here (locomotive unattributable and non-locomotive fuel costs) cannot be attributed to any particular shipper or group of shippers. Allowing a rail carrier to recover these incremental fuel costs—which represent only a small fraction of a rail carrier’s total incremental fuel costs—through a fuel surcharge mechanism does not present the sort of substantial unfairness to one set of shippers that motivated the Board to act in Fuel Surcharges.
Cargill’s reading of our cases is further weakened by the Board’s decision to specifically require that Class I rail carriers include locomotive unattributable and non-locomotive fuel costs in their Quarterly Reports to “enable the Board to better monitor the industry’s fuel surcharge practices.” Fuel Surcharges at 12. There would be little reason to require the inclusion of these particular incremental fuel costs in a report designed to allow the agency and the public to compare a rail carrier’s “[t]otal fuel cost” to its [t]otal revenue from fuel surcharges” if rail carriers were not permitted to recover these expenses in their fuel surcharge program.
The Board’s Fuel Cost Examination. The Board performed its own examination of BNSF’s month-to-month incremental fuel costs using the HDF Index. Because BNSF set its “strike price” at $1.25, we measured the change in BNSF’s fuel price above the $1.25 strike price by subtracting $1.25 from the monthly HDF price per gallon. And finding BNSF’s evidence on monthly gallons consumed reasonably developed and the best evidence of record, we accepted it. Thereafter, we multiplied the incremental fuel price for each month by the number of gallons of fuel BNSF consumed for each month and subtracted the resulting monthly incremental fuel costs from the monthly uncontested fuel surcharge revenues collected over the five-year period. On that basis, we find that over the five-year period BNSF’s fuel surcharge recovered 99% of total incremental fuel costs as measured by the HDF Index. We therefore conclude that Cargill has failed to show that BNSF’s fuel surcharge has been used to extract substantial profits above its incremental fuel costs for the BNSF system traffic to which the surcharge applied, and, as a result, that Cargill’s Profit Center claim lacks merit.
Before turning to the Reasonable Nexus claim, we note that we do not reject Cargill’s Profit Center claim lightly. The record here shows that, if measured by BNSF’s internal fuel costs (instead of the safe harbor HDF Index), BNSF’s fuel surcharge revenues exceeded its incremental fuel costs by some $181 million over the five-year period. We reject Cargill’s Profit Center claim, not because we think that what has turned out to be a $181 million recovery above BNSF’s internal fuel costs is insubstantial, but because, under the safe harbor provision adopted in Fuel Surcharges, rail carriers are entitled to rely on the HDF Index as a proxy to measure changes in their internal fuel costs. Using that index as the measure, BNSF has not over-recovered its incremental fuel costs and, as discussed in Part II, infra, no other aspect of its program has been shown to be unreasonable. We will seek public comment on whether changes to the safe harbor provision are warranted and will explain more fully in Part III our reasons for doing so.
II. The Reasonable Nexus Claim
Cargill claims that the general formula BNSF uses to calculate its fuel surcharge bears no reasonable nexus to, and overstates, fuel consumption for the BNSF system traffic to which the surcharge applies. Primarily, Cargill argues that BNSF’s fuel surcharge program is fundamentally flawed because it applies equally to agricultural and industrial products. In Cargill’s view, BNSF should have assumed different (and lower) fuel consumption rates for agricultural and industrial products, because these two types of traffic have significantly different fuel consumption characteristics, and should have adopted different fuel surcharges for them based on individually tailored step functions. Cargill also argues that BNSF should have adopted a higher HDF strike price of $1.298 per gallon and that the first 1¢ per mile incremental charge of the fuel surcharge should have begun to accrue at an HDF price of $1.324 for agricultural shipments and $1.321 for industrial shipments.
BNSF could have designed any number of fuel surcharges, each with its own step function, given the numerous traffic categories with different fuel consumption characteristics that could exist, depending on the distinctions that could apply to the over 5 million shipments that BNSF moved for Cargill and BNSF’s other shippers over the five-year period. But we do not find it unreasonable for BNSF to have adopted a single fuel surcharge equally applicable to agricultural and industrial products that is simple and easy for BNSF and its shippers to administer and use. Given that most of this traffic is transported at roughly similar fuel consumption levels, with much agricultural traffic transported along with other types of traffic in general merchandise trains, we cannot find this choice unreasonable. BNSF’s fuel surcharge program appears to be both straightforward and transparent. Thus, we do not find unreasonable BNSF’s decision to apply a single fuel surcharge equally to all agricultural and industrial products, even if it meant that the surcharge burden would not fall exactly equally on all types of traffic.
We also do not find anything unreasonable with BNSF’s assumption that a loaded rail car on average moves at a rate of 4 MPG. BNSF arrived at this assumption through what appears to be a careful process. According to BNSF, in June 2005, it examined the prior year’s data on traffic that would be subject to the mileage-based fuel surcharge, and, using highway miles, estimated that a loaded rail car on average moves at a rate of 3.92 MPG. BNSF states it then incorporated the 4 MPG Assumption into the proposed fuel surcharge, but to accommodate the concerns expressed by its agricultural and other carload customers, it decided to assess the proposed fuel surcharge using rail miles and delayed the implementation for non-agricultural shipments to give those customers time to adjust their internal systems. BNSF claims that it: (1) decided, consistent with the objectives of fairness and transparency, to use shortest rail miles as specified in its tariffs and not actual rail miles, a mileage measurement that tends to be longer; and (2) then reexamined the 2004 traffic data to determine whether the 4 MPG Assumption remained valid for agricultural traffic and, using actual rail miles this time, calculated that agricultural traffic moved at an average rate of 4.49 MPG.
In view of this process, BNSF reasonably retained the 4 MPG Assumption. For a mileage-based fuel surcharge to produce revenues in proportion to the rail carrier’s incremental fuel costs, it is important that the mileage source used to assess the fuel surcharge be the same as the mileage source uses to generate the MPG estimate incorporated into the fuel surcharge. Here, because BNSF decided to assess the proposed fuel surcharge based on shortest rail miles, it needed to incorporate an MPG assumption into its fuel surcharge calculated using shortest rail miles. BNSF had a 3.92 MPG estimate calculated using highway miles and a 4.49 MPG estimate for agricultural traffic calculated using actual rail miles. Because shortest rail miles usually fall between highway miles and actual rail miles, BNSF had reason to believe that an MPG estimate for agricultural and non-agricultural traffic combined, calculated on shortest rail miles, would fall somewhere between 3.92 MPG and 4.49 MPG. BNSF was justified in choosing a figure at the lower end of this range given that (1) if non-locomotive fuel had been included in the calculations, the range would have been about 0.2 MPG lower, BNSF Reply Statement, Anderson V.S. at 24, and (2) the 4.49 MPG estimate rested on an analysis of agricultural traffic only, even though the new fuel surcharge would eventually also apply to somewhat less fuel efficient non-agricultural traffic.
Cargill argues that BNSF’s internal MPG estimates undermine the 4 MPG Assumption. Cargill points to: (1) a 2005 study indicating that in 2003 and 2004 BNSF’s agricultural traffic averaged 4.89 and 4.86 MPG, respectively, based on actual rail miles, see Cargill Opening Statement at 28 & n.21; (2) fuel surcharge reviews from 2006 to 2010 showing that agricultural traffic averaged between 4.37 MPG and 4.47 MPG and that agricultural shuttle traffic averaged between 4.87 MPG and 5.02 MPG, id. at 29 n.24; and (3) a field study performed in 2006-2007 showing that agricultural shuttle trains achieved 5.65 MPG, id. at 30. None of these studies undermines BNSF’s decision to adopt the 4 MPG Assumption. The resulting MPG figures pertained to agricultural only or agricultural shuttle trains, and were calculated using actual rail miles, whereas BNSF required an MPG estimate for combined agricultural and industrial traffic based on shortest rail miles. Cargill’s own figures show that agricultural traffic is about 0.5 MPG more fuel efficient than industrial traffic, and that the use of actual rail miles rather than shortest rail miles to calculate fuel efficiency results in an apparent gain of approximately 0.5 MPG. See Cargill Rebuttal Statement 32-33. Moreover, none of the studies to which Cargill cites accounted for BNSF’s use of locomotive unattributable and non-locomotive fuel; if they had, the resulting MPG figures would have been still lower.
Nor was it unreasonable for BNSF to select a whole number and avoid using fractional amounts for its MPG Assumption. A whole number allowed BNSF to avoid giving the appearance that the estimate was more precise than it really was, which would help to avoid creating pressure for frequent changes in the fuel surcharge formula to account for relatively small changes in MPG estimates resulting from temporary fluctuations in traffic volumes or the mix of traffic subject to the fuel surcharge. Thus, a whole number was a reasonable choice to promote BNSF’s objective of simplicity and ease of administration.
Cargill’s challenge to the 4 MPG Assumption, as well as its challenge to the single, equally applied fuel surcharge program, if upheld, would require a far more complex fuel surcharge program. While a more complex program may also be reasonable, the assumptions BNSF used to achieve its desired simple and easy to administer approach are not unreasonable. Cargill’s regression analysis of the step function, on the other hand, is flawed in a number of key respects. Aside from ignoring the safe harbor HDF Index price of fuel, Cargill’s step functions are calculated using movement-specific adjustments to URCS variable costs, excluding locomotive unattributable and non-locomotive fuel costs, and adjusting actual internal fuel costs downward to account for hedging gains.
The choice of a $1.25 per gallon HDF strike price has also not been shown unreasonable. In general, rail carriers can recover fuel costs in a reasonable base rate, a reasonable fuel surcharge program, or a combination of the two. The choice of the strike price constitutes the rail carrier’s representation to the public of the delineation between the base rate and fuel surcharge. If a rail carrier sets the strike price at zero, it has represented to the public that it intends to recover its entire fuel costs in the fuel surcharge; if the strike price is set at infinity (i.e., it has no fuel surcharge program), the rail carrier has represented to the public that it will recover its entire fuel costs in the base rate. By choosing a strike price of $1.25, BNSF has informed the public that its base rates will cover all fuel costs where the HDF fuel price is less than $1.25 per gallon, and that incremental fuel costs above $1.25 per gallon will be recovered as a fuel surcharge. Our role in an unreasonable practice case is to assure that the rail carrier’s representation is not deceptive, rather than to order the rail carrier to make a different representation. In the end, we leave the decision of the strike price to the discretion of the rail carrier. So long as its representation to the public is not deceptive and absent evidence that the carrier has abused that discretion, we will not require a rail carrier to select a different strike point.
Here, we find neither deception nor abuse of discretion. As explained in Part I, BNSF’s representation that the fuel surcharge covers its increased fuel costs above $1.25 per HDF gallon is not deceptive. And BNSF has not abused its discretion in selecting the $1.25 per gallon strike price. The choice of the $1.25 strike price was a BNSF business decision that was based on a five-year historical average HDF cost and was intended to avoid further complications for BNSF’s shippers, who already had a few years of experience with a $1.25 strike price.
Based on the above findings and having previously confirmed that over the five-year period BNSF’s fuel surcharge recovered only 99% of the rail carrier’s total incremental fuel costs, as measured by the HDF Index, for the BNSF system traffic to which the surcharge applied, we cannot find that the general formula BNSF uses to calculate its fuel surcharges bears no reasonable nexus to the fuel consumption for the BNSF system traffic to which the surcharge applies. Accordingly, we find that Cargill’s Reasonable Nexus claim lacks merit.
III. The Safe Harbor
This proceeding has raised concerns about the safe harbor. For example, as noted at the end of Part I, the safe harbor can allow a rail carrier to recover more than its incremental fuel costs as measured by the carrier’s internal fuel costs. This can occur in those circumstances where changes in the HDF Index fail to accurately reflect changes in a rail carrier’s internal fuel costs, as happened here. See BNSF Reply Statement at 44-45 and Final Brief at 23 (“By 2010, a growing ‘spread’ had emerged between BNSF’s internal fuel price and the HDF index.”). Before the Board adopted the safe harbor provision in Fuel Surcharges, neither the Board nor any commenting party foresaw a situation where the spread between a rail carrier’s internal fuel costs and the HDF Index would grow divergently as happened here. We do not know if this was a unique situation affecting only BNSF during this period of high volatility in fuel prices or whether it is, or is likely to be, a more widespread phenomenon that may undermine the usefulness of the current safe harbor provision. Further, it has become apparent from this first fuel surcharge case to go forward that the safe harbor provision provides rail carriers with an unintended advantage. If a rail carrier’s internal fuel costs rise relative to HDF Index prices, the rail carrier can revise its fuel surcharge program to ensure that it fully recovers its incremental internal fuel costs. On the other hand, if a rail carrier’s internal fuel costs decline relative to HDF Index prices (as happened here during the five-year period), the rail carrier need not revise its fuel surcharge program. This would allow the rail carrier to recover more than its incremental internal fuel costs, and the safe harbor effectively would immunize that over-recovery from scrutiny.
While there is no evidence to suggest that BNSF intentionally took advantage of this aspect of the safe harbor, BNSF was aware of what was happening. Because it is possible that this aspect of the safe harbor provision could lead to future abuse, the Board will issue an advanced notice of proposed rulemaking to give shippers, rail carriers, and other interested parties the opportunity to comment on the safe harbor, including whether it should be modified or removed.
action will not significantly affect either the quality of the human
environment or the conservation of energy resources.
1. Cargill’s request to supplement its complaint is denied without prejudice.
2. This complaint is dismissed, and the proceeding is discontinued.
This decision is effective on the
By the Board, Chairman Elliott, Vice Chairman Begeman, and Commissioner Mulvey.
COMMISSIONER MULVEY, concurring with a separate expression:
I support the Board’s decision but write separately to express my concern about the Board’s suggestion that it will consider eliminating the use of the HDF Index as a safe harbor to measure changes in fuel costs. In Rail Fuel Surcharges, shippers argued “nearly unanimously” that the Board should adopt, in the words of Cargill, “a single, uniform index to measure fuel cost increases.” See Decision at 8. The Board designated the HDF Index as an index that rail carriers could rely safely upon as a transparent and widely accepted measure, but left the door open to any other indices that proved to be accurate, transparent, and neutral.
Cargill’s complaint against BNSF is the first fuel surcharge case to be fully adjudicated since Rail Fuel Surcharges was decided. This case has indeed raised difficult issues, due in large part to the now-evident increase in the spread between BNSF’s actual fuel costs and HDF Index costs at certain points during the 5-year period covered by Cargill’s complaint. Thus, it is appropriate for the Board to take another look at its fuel surcharge policy to see whether it needs to be modified, armed with the experience this case has given us. However, when the Board indicates that it will consider whether to “remove” the HDF Index as a safe harbor measurement tool, I think it goes far beyond any issues this case has revealed. Moreover, the implicit suggestion –even before receiving public comment – that it may not be appropriate for rail carriers to design fuel surcharge programs using the HDF Index could result in uncertainty and unintended consequences for stakeholders.
 The digest constitutes no part of the decision of the Board but has been prepared for the convenience of the reader. It may not be cited to or relied upon as precedent. Policy Statement on Plain Language in Decisions, EP 696 (STB served Sept. 2, 2010).
 Rail Fuel Surcharges (Fuel Surcharges—Proposed), EP 661 (STB served Aug. 3, 2006).
 See Rail Fuel Surcharges (Surcharge Reports), EP 661 (Sub-No. 1) (STB served Aug.14, 2007).
 The parties ultimately reached a settlement, and at Dairyland’s request, the Board dismissed the complaint with prejudice in a decision served on December 12, 2008.
 BNSF identifies the strike price as the “entry point” for its fuel surcharge. BNSF Comments at 16, Fuel Surcharges (Aug. 3, 2006). We have taken the strike price to mean the fuel price level at which the fuel surcharge begins to accrue.
In a letter filed on February 14, 2011, Cargill requested leave to supplement its complaint to include Item 3376D, Section B, as a successor to the challenged Item 3375L, Section B, and any other fuel surcharge tariff that BNSF might apply to common carrier shipments. In Item 3376D, Section B, of BNSF Rules Book 6100-A, BNSF rebased the strike price from $1.25 to $2.50 per gallon beginning January 2011, adjusted the mileage-based fuel surcharge for agricultural products effective on or around March 11, 2011, and added a credit mechanism to compensate customers if the average HDF price fell below the $2.50 strike price. BNSF, in a letter-reply filed on February 22, 2011, argued that Cargill should not be permitted to amend or supplement its complaint, asserting that Item 3376D, Section B, sets out the terms of a new fuel surcharge program, which under 49 C.F.R. § 1111.2 requires the filing of a new complaint. Cargill filed a letter-reply to BNSF’s reply and a pleading styled “Supplement to Complaint” on March 4, 2011, and BNSF filed an “Answer to Supplement to Complaint” on March 22, 2011. Because Item 3376D, Section B, became effective not long before the close of discovery in this proceeding, the record lacks evidence concerning its operation. Accordingly, we deny without prejudice Cargill’s request to supplement its complaint. If it chooses, Cargill may challenge Item 3376D, Section B, or any subsequent iteration, by filing another complaint.
 In its Opening Statement at 40 Cargill sought in excess of $29 million in damages plus interest; in its Rebuttal Statement at 6 it seeks $26.8 million in damages plus interest.
 In Union Pacific, 867 F.2d at 649, the D.C. Circuit concluded that our predecessor, the Interstate Commerce Commission, had impermissibly engaged in rate regulation when it found that charging more to offset extra costs incurred in transporting spent nuclear fuel constituted an unreasonable practice. The Board’s pronouncements in Fuel Surcharges were designed to conform to that holding.
 See Cargill, Inc. v. BNSF Ry. (2011 Decision), NOR 42120 (STB served Jan. 4, 2011).
 See Cargill, Inc. v. BNSF Ry., NOR 42120 (STB served May 25, 2012).
 “Incremental fuel costs” are those fuel costs not embedded in the base rate and which a rail carrier seeks to recover through a fuel surcharge mechanism. As explained later in this decision, Cargill and BNSF differ significantly on the proper way to measure incremental fuel costs.
 Having concluded that Cargill has not shown that BNSF over-recovered its incremental fuel costs, as properly measured by the safe harbor HDF Index, we need not address such issues as the level of over-recovery that would be required to establish an unreasonable practice, the point in time that a rail carrier’s over-recovery of incremental fuel costs becomes an unreasonable practice, and whether an unreasonable practice could be established by a showing of an over-recovery without a showing that the fuel surcharge lacks a reasonable nexus to fuel consumption.
 The parties designated this information as “confidential.” Although the Board attempts to avoid references to confidential or highly confidential information in our decisions, we reserve the right to rely on and disclose such information when necessary. In this case, we have determined that we could not adequately present our findings with respect to the Profit Center and Reasonable Nexus claims without disclosing confidential information here and elsewhere in this decision.
 URCS, the Board’s general purpose costing system for all regulatory costing purposes, is used to determine a carrier’s variable costs in a variety of regulatory proceedings. See Adoption of the Unif. R.R. Costing Sys. as a Gen. Purpose Costing Sys. for All Regulatory Costing Purposes, 5 I.C.C.2d 894 (1989).
 In its comments, Cargill responded affirmatively to the Board’s proposal, stating that “the STB needs to establish a single, uniform index to measure fuel cost increases.” Cargill Comments at 4, Fuel Surcharges (Oct. 2, 2006). Cargill supported a single uniform index but suggested an alternative to the HDF Index.
 In the 2011 Decision at 5, the Board stated as follows: “Consistent with Dairyland, Cargill may present evidence to demonstrate that design elements in the challenged fuel surcharge allow BNSF to recover substantially in excess of the actual incremental cost of fuel incurred in providing the rail services to the entire traffic group to which the surcharge applies.”
 In this context, the Board’s use of the term “actual incremental fuel costs”—which was broader than needed—cannot be construed to implicitly overrule the safe harbor.
 Indeed, Cargill’s expert arrived at its “consumption adjustment factor” based, not on any information available to BNSF before it instituted the challenged fuel surcharge, but rather on after-the-fact data from 2006 to 2010. BNSF literally had no way of using Cargill’s proposed consumption adjustment factor in developing its fuel surcharge formula.
 See Major Issues in Rail Rate Cases, EP 657, slip op. at 60 (STB served Oct. 30, 2006).
 See Entergy Ark., Inc. v. Union Pac. R.R., NOR 42104, slip op. at 13 n.37 (STB served Mar. 15, 2011) (rejecting movement-specific adjustments in a competitive access case under 49 U.S.C. § 10705).
 Of course, shippers who wish to be subject to the potential risks and rewards of hedging are free to engage in their own hedging activities, and some, such as Cargill, do. See Cargill Comments at 2-3, Fuel Surcharges (Oct. 2, 2006) (“[A]s a regular participant in the commodities markets, Cargill has developed skills for managing cost/price volatility. Cargill employs those skills to hedge against the changes in fuel costs that it pays to carriers through fuel surcharges.”).
 See Report of Fuel Cost, Consumption, and Surcharge Revenue, 49 C.F.R. § 1243.3 (2012).
 BNSF had performed an incremental fuel cost study using the HDF Index to measure the monthly changes in its fuel costs. To estimate monthly total gallons of fuel consumed, BNSF: (1) used URCS to develop an estimate of the variable locomotive fuel costs associated with each of the over 5 million individual shipments subject to the challenged fuel surcharge that it moved for Cargill and other shippers over the five-year period; (2) increased that estimate by various percentages derived from its Quarterly Reports to account for locomotive unattributable and non-locomotive fuel consumption; and (3) divided the resulting total fuel expense estimate by the system-average BNSF fuel price per gallon implicit in URCS. According to BNSF’s HDF/safe harbor cost study, it recovered 99% of the total incremental fuel costs that it incurred over the five-year period.
 Similarly, BNSF’s decision to set the initial penny charge of the surcharge at an HDF price of $1.25 per gallon, rather than at the midpoint of the first step of the fuel surcharge step function, has not been shown unreasonable. The Board requires only that a fuel surcharge program have a reasonable nexus to fuel consumption; we will not micromanage the details of a rail carrier’s fuel surcharge program.