UNITED STATES COURT OF APPEALS ) FOR THE EIGHTH CIRCUIT ) CITY OF BRIDGETON, MISSOURI, et al., ) Petitioners, ) ) Petitions for Review v. ) Nos. 98-3506; 98-3774; 98-3925 ) RODNEY E. SLATER, et al., ) Respondents. )
Richard P. Sher
Stephen C. Murphy
DEVEREUX, MURPHY, STRILER,
BRICKEY & SHER, L.L.C.
10 S. Brentwood, Suite 215
St. Louis, MO 63105
Attorneys for Amicus Curiae
National Air Traffic Controllers Association, Local T-75
Table of Contents
Table of Authorities
Identity of Amicus Curiae
(1) The W-1W Plan Will Create a More Complicated, Inefficient Airport Environment, with Excessively High Controller and Pilot Workload
(2) The W-1W Plan Does Not Increase Capacity at Lambert-St. Louis International Airport in Any Significant Way; As Air Traffic Increases In The Future, and W-1W Performance Does Not Adequately Handle That Traffic, Pressure Will Build On Controllers and Pilots to Reduce the Margins of Safety in an Effort to Reach the Levels of Performance Promised in the Master-Planning Process
(3) The FAA's Analysis in Support of W-1W Is Flawed in Numerous Respects, Rendering its Decision Approving the Plan Unreasonable
(4) The Failure of the FAA to Test the Capacity of W-1W by Conducting a Real-Time Simulation, Under Appropriate Assumed Conditions, Renders its Decision Arbitrary and Capricious
49 U.S.C. § 40101(a)(1) 2
42 U.S.C. § 4332(C) 11
40 C.R.F. § 1502.14 (1996) 11
Dubois v. U.S. Dp't of Agriculture, 102 F.3d 1273 (1st Cir. 1986), cert. denied, 117 S. Ct. 2510 (1997);
Friends of Boundary Waters Wilderness v. Dombeck, No. 97-3282, 1999 U.S. App. LEXIS 364*36 (8th Cir. Jan 7, 1999).
The Air Traffic Controllers Association, Local T-75 ("NATCA"), is comprised of the air traffic controllers who are responsible for directing the approach and departure of aircraft around Lambert-St. Louis International Airport. The members of NATCA will be directly responsible for implementing the air-traffic requirements of the expansion plan for Lambert, and thus have a direct interest in the process by which a plan is selected.
Pursuant to Rule 29 of the Federal Rules of Appellate Procedure, National Air Traffic Controllers Association, Local T-75("NATCA") has sought leave to file this brief as amicus curiae in support of the petitions for review. This brief sets out the position of NATCA's members regarding the order of the Federal Aviation Administration ("FAA") entitled "Record of Decision, Lambert-St. Louis International Airport, September 30, 1998," entered on September 30, 1998.
The function of air traffic controllers is to provide for the safe, orderly and expeditious flow of aircraft in and around airports. Controllers are responsible for directing the approach and departure of aircraft, providing sequencing of aircraft, and for separating arrival and departure traffic around the airport. Members of the National Air Traffic Controllers Association, Local T-75 ("NATCA") are employees of the FAA.
NATCA opposes the airport expansion plan (known as W-1W) approved by the FAA in this case. Its opposition is unprecedented. Generally, any improvement in capacity, technology, or procedure is welcomed by controllers. Approval by the FAA of the W-1W plan, however, was made without meaningful input by the air traffic controllers (and the pilots) who will be required to implement the plan and in disregard of fundamental safety issues. The plan was also approved without a "real-time simulation," the most realistic test upon which the FAA's assumptions with regard to increased traffic flow at Lambert-St. Louis International Airport could be properly based. NATCA believes the FAA's approval of W-1W was unreasonable for the following reasons:
(2) the W-1W plan does not increase capacity at Lambert-St. Louis International in any significant way; as air traffic increases, and the airport, as expanded, fails to adequately handle that traffic, pressure will build on controllers and pilots to reduce the margins of safety in an effort to reach the levels of performance promised in the FAA's master planning process;
(3) the FAA's analysis in support of W-1W is flawed in numerous respects; and
(4) the FAA failed to test actual capacity under the W-1W plan by conducting a real-time simulation.
Each of these concerns was communicated by controllers and pilots to the FAA leadership, but the review process was conducted so as to favor political considerations over practical and safety issues. The result, the W-1W plan, is an extremely expensive, technically inferior plan that provides little of what has been promised to the public.
NATCA's primary goal (in relation to W-1W) is to see the National Airspace System (NAS) expanded properly. Because the greatest impediment to the continued growth of air traffic in the United States is the lack of adequate airport capacity, air traffic controllers are staunch supporters of plans that meaningfully improve airport capacity. But, after extensive review of the W-1W plan, NATCA firmly believes that W-1W is expensive, impractical, cumbersome to operate, and offers little improvement to current arrival capacity.
Under 49 U.S.C. §40101(a)(1), "assigning and maintaining safety is the highest priority in air commerce." In providing air traffic control services, NATCA members will always comply with the safety standards set forth in the applicable FAA manuals and directives. If W-1W is operated in compliance with these FAA rules and directives, however, there will be little meaningful capacity improvement; instead, there will be significantly increased controller and pilot workload, making safety compliance more difficult.
The unique configuration of the W-1W plan, requiring pairing of aircraft with different approach speeds, combined with different final speed reduction points along the approach path, are major problems for air traffic control. Traffic flows must also be manipulated by controllers to accommodate the imbalance caused by the unusual spacing requirements necessary to run this configuration. The arrival acceptance rates to runway 30W would be half that of arrivals on runway 30R. This creates unnecessary stress on the "feeder" approach control position, especially during high demand from the west and southeast inbound routes. The complexity of the proposal is unlike any other in the United States, and its use has therefore never been tested except in theoretical, computerized simulations.
The FAA represents that W-1W will significantly increase capacity at Lambert-St. Louis. However, this representation is premised on both an understatement of present capacity and overstatement of projected capacity under W-1W -- both based on computer models rather than experience with a simulation under real conditions.
The problem with the W-1W Plan is that it does not accomplish the FAA's stated and intended goal of achieving independent and simultaneous arrivals of aircraft on two runways. W-1W does not properly take into account the impact that these arrival operations would have on departures. This is particularly true when the runways are used in west flow operations (when planes arrive and depart to the west), which presently comprise 60 percent of all operations at Lambert. Since normally there are both arrivals and departures occurring at the same time, how arrival runways impact departure runways must be considered. Without any departures, the two outboard runways can run independent arrivals in instrument conditions. However, when departures are added to the equation, the benefits of the independent arrival runways are largely lost. This is due to the long threshold stagger on the landing runway (30W) and the dependency it creates on departures off the longest departure runway (30L). To comply with air traffic rules, a large arrival gap must be maintained on runway 30W to allow for a runway 30L departure. In addition, the departures off this same runway would remain dependent on the arrivals on the present runway 30R. Since this scenario creates a double dependency, arrivals must be operated in pairs to insure that the dependencies are resolved simultaneously. Normal instrument operations in this configuration, therefore, are not independent.
NATCA believes that the unusual geometry of the W-1W configuration renders the plan operationally cumbersome and creates a high controller workload. Controllers and pilots agree that W-1W will be more difficult to operate safely. No other airport in the United States has the combination of proposed geometry and suggested air carrier runway use that W-1W presents. Airport planners simply do notdesign airports that look and operate like W-1W.
Because of the staggered runway thresholds, the W-1W design creates arrival and departure dependencies that are not operationally acceptable. The FAA's own analysis (and disapproval) of these dependency problems appeared in the Record of Decision (ROD) for the airport expansion in Seattle. The Seattle ROD described any runway stagger that created dependencies in the predominant flow as "inappropriate." In considering the possibility of a single runway dependency brought about by threshold staggers at Seattle (W-1W will have a double dependency), the FAA noted, "For operational reasons, the stagger option is not practical. Where a stagger exists greater coordination is required between arrival streams and departures. . .a staggered threshold would produce inefficiencies." The FAA also remarked: "This added complexity in air traffic operations would occur at times when the operations are at their highest level and during poor weather, a time when air traffic controllers are at their highest work load." NATCA agrees with this assessment.
The FAA Technical Center's "Lambert-St. Louis International Airport Capacity/Delay Study," in August, 1992, for a proposed airport expansion with a nearly identical physical layout as
W-1W indicates that W-1W delays were understated in the Final Environmental Impact Statement (FEIS) by nearly a factor of three, despite the fact that, in 1992, the FAA used 50,000 fewer operations than it did in the FEIS to arrive at its delay number -- use of increased operations would cause the FAA claims on W-1Wdelays to be even further from actuality. Moreover, the FAA's Advisory Circular on Airport Capacity and Delay, 150-5060-5, clearly indicates that W-1W delays as presented in the FAA's Master Plan Supplement ("MPS") are grossly understated.
The "SIMMOD Review," a computer simulation conducted by the FAA Technical Center in 1998, concluded that an additional 1.5 minutes of average delay should be added to W-1W to account for underestimated taxi times in the Master Plan Supplement (MPS) and FEIS. NATCA believes that even this revised number is understated, because the FAA claimed that runway crossing time ended when the aircraft tail exited the runway. Such an aircraft position is not considered to be across the runway. NATCA must insure that the aircraft is clear of the runway safety area, about another 175 feet beyond the runway edge, before a runway crossing is complete.
NATCA informed the FAA that the W-1W use of the present runways 30R and 30L for arrivals in good weather conditions would produce only about 60 arrivals per hour--a number supported by the Airline Pilot Association (ALPA) and FAA tower management. Based on the assumption of 60 arrivals per hour for W-1W (instead of the 72 that were erroneously assumed in the MPS and FEIS), NATCA, at a St. Louis Airport Authority briefing to the FAA, produced a chart which showed W-1W average delays to be 8.6 minutes per aircraft. This delay, added to the 1.5 minutes of taxi time addressed above, results in an average delay per aircraft for W-1W of 10.1 minutes -- almost triple the amount the FAA considers an acceptable delay. This is more than a 100% increase (10.1 minutes vs. 4.7 minutes) in the delays for W-1W set forth in the MPS.
The SIMMOD Review also erroneously stated that the calculated average delay per aircraft for the W-1W Plan was valid because similar delays were calculated in the FAA 1988 Capacity Enhancement Study for St. Louis. Analysis of the FAA 1988 study, however, indicates that the runway geometry used was similar to the S-1 Plan, not to W-1W, as claimed. Furthermore, the average delay per aircraft in the FAA 1988 study was derived from a proposed procedure that assumed arrivals on three runways at St.Louis in poor weather. Such a procedure has never been implemented at St. Louis. The attempt to validate the W-1W SIMMOD results based on correlation with unused procedures on a different runway configuration is totally invalid.
In visual flight (good weather) conditions, arrivals will be directed to the existing runways under the W-1W Plan. The fact that these arrivals are dependent limits arrival capacity to approximately 60 aircraft per hour. The FEIS assumes that these are independent arrivals, and assumes 72 arrivals per hour, seriously overstating the arrival capacity. Moreover, this use of existing runways will not increase Lambert's absolute arrival capacity. Indeed, it may reduce capacity because of conflicts between arrivals on runway 24 and aircraft taxiing to depart on runway W-1W.
The use of the W-1W runway in a west flow (60% of the time) is not independent, as claimed in the MPS and FEIS. In poor weather, W-1W arrivals on parallel runways are "paired" which is a dependent arrival operation. In good and marginal weather, W-1W arrivals are conducted on parallel runways that are spaced 1300 feet apart -- also a dependent arrival operation. Dependent operations, particularly in the case of poor weather, are highly complex, labor-intensive practices. For this reason they are important factors in airport delays, as pointed out in the ROD for the Seattle Airport expansion.
The process that resulted in the selection of W-1W was flawed in the following respects: (a) it underestimated the capacity of the present airport; (b) it overestimated the capacity of the W-1W plan; (c) it inadequately considered alternatives to W-1W; and (d) it excluded meaningful input from air traffic controllers.
(a) Present Capacity of Lambert. One of NATCA's strongest objections with the master planning process involves the lack of landing capability attributed to the "Converging Runway Display Aid" (CRDA) currently being used at Lambert. The Record of Decision incorrectly states that "the airport is reduced to one precision instrument approach in adverse weather conditions because of the minimal separation of the parallel runways." Historical data and charts within the FEIS clearly show that the present airport has the capability of using more than one approach. This premise -- lack of two poor weather arrival runways -- is used to support the MPS contention that providing approach capability to two runways in poor weather at Lambert is a significant step forward relative to the present airport. This is simply not the case.
Lambert was the national test site for CRDA, a tool that utilizes a computer enhancement to provide controllers with the ability to land on converging runways in all weather conditions at Lambert. The use of instrument approaches on runways 30L and 30R in marginal weather conditions ("LDA") also enhances capacity levels. In the MPS, FEIS, and St. Louis Airport Authority (STLAA) briefings to the FAA on July 20th, 1998, FAA management asserted that landing capacity with CRDA was 38 per hour, and the City's airport consultant asserted it was 40 per hour. NATCA, by actual airport arrival count, documented by videotape, determined that CRDA can land as many as 60 arrivals per hour. To account for any number of variables, NATCA conservatively put the arrival rates for CRDA at 48 to 54 per hour.
Any number of documents support the NATCA CRDA arrival rate claim, including an FAA Central Region Press Release, an article in the FAA Central Region Magazine, and a technical discussion of CRDA authored by MITRE, the FAA contractor who developed CRDA. Considering the fact that a single runway at St. Louis today can arrive 36 aircraft per hour, the FAA management number of 38 arrivals for two runways is clearly erroneous. It is inconceivable that the FAA would spend millions of dollars to develop and test a CRDA-based procedure that only produced two additional arrivals per hour.
The airport consultant's conclusion of 40 arrivals per hour for CRDA in the existing Lambert configuration was generated by a computer simulation, when simply counting the arrivals would have yielded a far more accurate indication of CRDA capability. "Simulating" CRDA makes no more sense than would a computer simulation of automobile flows on Interstate 70 in St. Louis when an exact traffic count could be obtained by simply counting the cars. The inescapable conclusion is that the actual, present CRDA capacity was misstated in the MPS so as to intentionally downgrade the capability of the present airport and make the W-1W plan appear relatively better.