Chicago, IL, USA
B-18711
BOEING 747
CAL5148 was conducting an ILS autoland approach to runway 10R. Instrument meteorological conditions prevailed at the time of the incident, with an overcast ceiling of 600 feet above ground level (agl), and visibility of 2 and 1/2 miles with moderate rain. These conditions required protection of the ILS critical area by ATC which they did according to FAA Order JO 7110.65, Air Traffic Control. ATC issues control instructions to protect ILS critical areas and avoid signal interference from vehicle traffic and taxiing aircraft when the reported weather is a ceiling of less than 800 feet or visibility of less than 2 miles. However, this restriction does not include consideration for preceding arriving or departing aircraft. While a previously departed aircraft was airborne and over the ILS critical area, CAL5148 experienced localizer deviations while on short final. In math modelling analysis of this event conducted by the FAA; the results were consistent with the localizer deviations determined from CAL5148 flight data recorder (FDR) data. Specifically, the modelling indicated the deviation beginning at 1/4 mile from threshold would lead the aircraft to turn left to continue to follow the localizer course line during an autoland approach. The FAA National Resource Engineer for Navigation who performed the modelling wrote “I am 90% confident that the location of departing ENY3544 caused sufficient Multipath (signal deflection) to produce a fly left indication to arriving CAL5148.” “Multipath” is a term used to describe the distortion of the radiated signal which is caused by the signal being deflected off either stationary or moving objects such as aircraft or vehicles in the localizer or glide slope critical area(s).
HISTORY OF FLIGHTCAL5148 was performing an Instrument Landing System (ILS) autoland approach under conditions that allowed for the exception listed in FAA Order JO 7110.65, Air Traffic Control paragraph 3-7-5a1(a)(2) to be applied. Envoy flight 3544 (ENY3544) had departed runway 10L but was still over the ILS critical area while CAL5148 was attempting to land. Audio recordings of ATC communications were provided by the FAA and were used in producing the following timeline. Times are in CDT and rounded to the nearest minute, and altitudes are in feet above mean sea level (msl). 0956 The KORD ATCT local control (LC) controller instructed ENY3544 to line up and wait on runway 10L. At the same time, CAL5148 was on a 4-mile final for the same runway. 0957 CAL5148 reported inbound on the ILS runway 10L and was instructed by the LC controller to continue inbound and expect clearance in approximately 2 miles. 0957 ENY3544 was cleared for takeoff by LC controller. 0958 The LC controller cleared CAL5148 to land runway 10L and advised the pilot of traffic [ENY3544] departing the same runway. 0958 CAL5148 crossed landing threshold at about 700 feet msl (less than 100 feet above the ground). 0959 The first officer of DAL2012 holding short of runway 10L at taxiway DD stated “goin off the runway.” 0959 CAL5148 reported going around due to “last minute deviation on the runway.” While on short final and attempting to land, CAL5148 drifted left of the runway centerline and the left main landing gear touched down left of the runway, but still on the paved surface. The airplane then traveled into the grass between taxiway DD and taxiway N1 leaving deep ground scarring and damaging a windsock. Figure 1 shows the accident airplanes flight track overlaid on satellite imagery in which the ground scarring can still be seen. CAL5148 executed a go-around and subsequently returned to the airport, landing on runway 9R without further incident at about 1020. Figure 1. Figure showing a portion of CAL5148’s flight track, with initial touchdown area and other pertinent information also illustrated. METEOROLOGICAL INFORMATIONKORD had an Automated Surface Observing System (ASOS) whose reports were supplemented by certified weather observers. The ASOS site was located at an elevation of 680 feet msl and had a 3° westerly magnetic variation. Surface Observations [0934 CDT] SPECI KORD 211434Z 06010KT 2SM RA BR OVC006 19/18 A2978 RMK AO2 LTG DSNT E P0019 T01890178= [0951 CDT] METAR KORD 211451Z 06012KT 2 1/2SM RA BR OVC006 19/18 A2978 RMK AO2 LTG DSNT E SLP080 P0025 60056 T01890178 58001= KORD weather at 0951 CDT, wind from 060° at 12 knots, 2 and a half statute miles visibility, moderate rain, mist, overcast ceiling at 600 feet agl, temperature 19° Celsius (C), dew point temperature 18° C, altimeter 29.78 inHg. Remarks, station with a precipitation discriminator, lightning distant to the east, sea level pressure 1008.0 hectoPascals (hPa), 0.25 inches of precipitation since 0851 CDT, 6-hourly precipitation of 0.56 inches, temperature 18.9° C, dew point temperature 17.8° C, 3-hourly pressure decrease of 0.1 hPa. Terminal Aerodrome Forecast Terminal Aerodrome Forecasts (TAF) were issued by the NWS for KORD and the KORD TAF that was valid at the time of the incident was issued at 0812 CDT and is provided below: TAF KORD 2113/2218 07010KT 3SM -SHRA BR SCT005 OVC009 TEMPO 2113/2115 3/4SM +RA BR OVC005 FM 211500 08012G17KT 3SM RA BR VCTS SCT009 BKN015 OVC025CB TEMPO 2115/2119 1 1/2SM +RA BR BKN009 OVC015 FM 212000 07012G17KT 5SM -SHRA BR SCT009 OVC012 FM 220100 07012KT 6SM -SHRA BKN010 OVC020 FM 220600 05012KT 5SM RA BR SCT007 OVC012 Between 0800 and 1000 CDT (on June 21), the forecast expected a wind from 070° at 10 knots, 3 miles visibility, light rain showers and mist, scattered clouds at 500 feet agl and an overcast ceiling of 900 feet agl. Temporary conditions were forecast between 0800 and 1000 CDT of 3/4 miles visibility, heavy rain, and an overcast ceiling of 500 feet agl. WRECKAGE AND IMPACT INFORMATIONDeep ground scarring was visible near where the aircraft had touched down left of the runway on the paved surface and extending into the grass between taxiway DD and taxiway N1 (see figure 3). Figure 3. Photographs of the ground scarring caused by the left main landing gear of CAL5148. [Source: FAA] FLIGHT RECORDERSThe accident airplane was equipped with a Honeywell 4700 solid-state FDR that contained about 27 hours of flight data. The accident flight was the last flight of the recording, and its duration was about 5 hours and 52 minutes. The data were extracted normally. The airplane was also equipped with a Honeywell CVR capable of recording 120 minutes of digital audio. It contained a two-channel recording of the last 120 minutes that power remained applied to the aircraft, however because power remained to the aircraft long after the event, all audio information from the accident flight was overwritten. TESTS AND RESEARCHMath modelling of this event was conducted by the FAA’s National Resource Engineer for Navigation. The FAA applied a model that could consider if objects, like buildings and aircraft, could affect localizer and glide slope quality; however, this model was not capable of assessing the effects of an aircraft flying directly over the antenna. Therefore, an analysis was performed with ENY3544 located 2,800 feet from the localizer and 450 feet above the runway. Using several assumptions for aircraft orientation, the model’s output was consistent with the localizer deviations determined from the CAL5148 FDR data (see figure 4). Figure 4. Graph displaying the simulation results of the math modelling conducted by the FAA. The graph in figure 4 illustrates modelling simulation results of the effect of a regional jet [ENY3544] departing and overflying the runway 10L localizer at KORD. The location of the Regional Jet simulated was based on Airport Surface Detection Equipment – Model X (ASDE-X) surveillance data of the regional jet and B747 [CAL5148] positions around the time of the runway excursion. On the graph, the red line represents predicted deviations of the aircraft course deviation indicator for the localizer. The vertical scale is the predicted deflection in micro-amps. For an aircraft course deviation indicator, 150 micro-amps would produce full-scale deflection of the indicator. The horizontal scale is the B747 position referenced to runway threshold. The distance labels “0” and “T” are at the threshold. The numerical values are in nautical miles. The letters are used to illustrate flight inspection areas of interest. Point “D” is 3,000 feet down the runway from the threshold and is the end of the touchdown zone. The deviation beginning 0.25 mile from threshold would lead the aircraft to turn left to continue to follow the localizer course line during an autoland approach. Upon completion, the National Resource Engineer for Navigation who performed the modelling wrote, “I am 90% confident that the location of departing ENY3544 caused sufficient Multipath (signal reflection) to produce a fly left indication to arriving CAL5148.”
Interference with the localizer signal resulting in a Multipath condition experienced by CAL5148 while conducting an ILS CAT III (autoland) approach. This condition resulted in a runway excursion and subsequent go-around. The interference was most likely caused by a previous departure flying through the ILS critical area while departing the same runway.
Source: NTSB Aviation Accident Database
Aviation Accidents App
In-Depth Access to Aviation Accident Reports
