How low Did You Go?

How low Did You Go?

How low Did You Go?

The Bell 212 crew had waited in vain for seven hours, hoping the weather would improve at the departure airport, Alerk Island, NWT, and the destination, an oil rig in the Beaufort Sea. The helicopter was finally dispatched in night IFR conditions to fulfill the leasing company’s requirement to have a machine and crew located at the drill site.

Once in range of the rig, the crew commenced an improvised IFR approach, in ice fog, descending to 500 ft MSL, an altitude believed to assure sufficient separation from the platform and sea below. The weather at the time was: wind east at 10 knots, visibility 1 ½ in fog, 400 ft overcast, temperature -19C.

kbtv101wapt.pdfAs the platform passed beneath them, the crew saw the glow from the rig and proceeded briefly outbound, then turned back in the direction of the lights, using a tear-drop maneuver while descending to a target altitude of 150 ft for the inbound leg. Slightly after reaching that altitude, and without warning, the helicopter struck the sea ice. It disintegrated upon impact; a post-impact fire completely consumed the aircraft. One of the two crew members suffered serious injuries.

There are a number of troubling aspects to the way this flight was conducted, including improper in-flight decision making, improper IFR procedures, a poorly planned (improvised) approach, and failure to go missed once visual reference was lost. But the Canadian Safety Board listed two primary causes in its report that speaks to the inherent dangers present when subtle errors creep into the aircraft’s altimetry. In the case of this Bell 212, the pilot’s altimeter, in compliance with regulations, read 50 feet low when set with the proper barometric pressure. Still, it was the crew’s failure to accomplish the proper temperature correction that likely doomed the flight.

To understand how this could happen, it is useful to review the recently-updated FAA-issued temperature correction chart. While temperature correction hasn’t been mandatory in the United States for designated airports until recently, it has been understood for a long time that very low (below ISA) temperatures can have a dramatic effect on pressure altimeter accuracy. Knowing the temperature measured by the rig operators was -19C, at 150 feet the crew should have added 30 feet to their target altitude. Not doing so put the aircraft only 80 feet above the ice when the altimeter read 150 feet. With very little visual reference, and while searching in the morass for the oil platform, it’s easy to see how whatever buffer remained was consumed.

kbtv124app.pdfThere are now 284 airports in the United States where temperature correction is required to complete one or more instrument approaches. Each airport is listed showing the segment that is affected, Intermediate (after the IAF/Intermediate fix and prior to and including the FAF), Final (after FAF to MDA or DH), and/or Missed (all published altitudes), and the aerodrome temperature governing when temperature correction must be accomplished. For some airports, temperature correction is required for multiple segments to complete the approach safely and legally. When there are multiple segments identified, each one may have its own low-temperature trigger point.

So how does the pilot know whether a particular airport is impacted and to which segment(s) the calculation should be applied? The FAA’s list of affected airports and applicable segments is published in the General NOTAMs section, along with aerodrome temperature that must be reached before temperature corrections apply.

In the case of the Jeppesen charts, the Cold Temperature Correction
Table is included as part of the plate packet, with the note section of each individual approach plate showing the maximum temperature not requiring correction. The government publications show the same information in the AIM 7-2-3, ICAO Cold Temperature Error Correction Table and as a snowflake symbol on the plate, respectively.

kbtv112app.pdfOnce it’s determined that temperature correction is required and the applicable segments are identified, the pilot must calculate the proper adjustment. Our CJ3’s Rockwell Collins Proline 21 suite has a special setting that allows the computer to automatically calculate the temperature correction for each affected waypoint, using the aerodrome temperature for input. For airplanes not so equipped, the pilot must make the calculation manually. In either case, it’s the pilot’s responsibility to inform ATC that the approach will be flown using temperature-corrected altitudes. And while it is illegal (and unsafe) to fly an approach requiring temperature correction without making the necessary corrections, ATC is under no obligation to inform the pilot as to the transgression. Ascertaining applicability and making the proper correction is 100% a pilot responsibility.

Actually doing the calculation is relatively simple, but there are some pitfalls to be avoided. First, the correction table is based on a zero elevation relative to the Above Airport Ground Level (AAGL) height of the waypoint where the calculation is being made. So, for example, in cases where the prototypical Burlington, Vermont’s ILS/LOC Rwy 33 approach requires temperature correction, the airport elevation (335’) must be subtracted from the MSL waypoint altitude before referencing the correction table. Assuming a -30C day, typical in Vermont mid-winter, and knowing from the NOTAM that only the intermediate segment is affected:

Screen Shot 2016-03-07 at 8.38.18 AM* rounded up from the table calculation for purposes of better altitude targeting in the cockpit and clearer communication to ATC.

It’s important that the pilot not make an altimeter-setting change to accomplish temperature correction. Also, although the correction value is always calculated by interpolation or rounding up (not down) from the closest table AAGL altitude, as in this example, extrapolation for AAGL altitudes greater than 5000’ is not required. For AAGL altitudes over 5000’, the 5000’ column is used.

The Burlington case also shows that altitude corrections tend to increase in magnitude the higher the waypoint is above the aerodrome altitude. At the airport, there is, by definition, no temperature-related error, because barometric pressure and temperature sampled from the same place largely correlate with the correct altitude (as the temperature goes down, the pressure increases). But, at altitude, the temperatures are often lower than at the airport below. If they are low enough below ISA, the airport’s barometric pressure that is used to set the airplane’s altimeter and the pressure at altitude measured by the airplane’s altimeter don’t correlate well with the actual MSL altitude. If the temperatures are low enough, the error can be very large. Proof by example, the government’s formula assumes a –2C degree/1000’ normal air lapse rate. In cases where the actual lapse rate is more, the error can be even worse than the temperature correction table predicts. We see this in Vermont occasionally, following the passing of a strong cold front.

The admittedly conservative Burlington example should give pause because, without correction, an airplane flying at the published minimum altitude of 4800’ MSL between Niduq and Honib could actually be as low as 3850’MSL on this particular day; 176 feet below the top of Camel’s Hump Mountain if tracking just slightly left of the highest point (4026’ MSL). Also, consider that the approach plate specifically forbids autopilot-coupled approaches, making precise adherence to vertical and lateral guidance just a little more challenging. Even on a somewhat warmer day, the buffer could be perilously thin. Scary!

RNAV approaches are much like ILS and LOC approaches, but it is important to understand when baro altimetry-related temperature correction applies. LPV approaches follow the same rules appropriate to ILS approaches because the glide slope is 100% determined by GPS WAAS satellite information. In contrast, LNAV approaches inherently dependent on barometric altimetry are subject to temperature correction, much like LOC or VOR approaches whenever the airport’s Final segment is called out within the NOTAM document and aerodrome temperatures apply. Temperature correction for baro-dependent approaches should not be confused with low temperature restrictions that apply to LNAV/VNAV approaches at certain airports. For example, in the case of Burlington’s RNAV Z (GPS) approach to 33, the baro-derived LNAV/VNAV approach is forbidden for airplanes not equipped with automatic temperature compensation  (to adjust the glide slope) when the airport temperature falls below -15C.

It is curious that there haven’t been more accidents related to low temperature altimeter error. Part of the reason may be that low temperature days in the USA tend to be clear days; mostly visual approaches. There is also a lot of buffer built into the design of instrument approaches to mitigate for other sources of mistakes, including less-than-precise navigation and instrument error. ATC also deserves some credit in the sense that, except for a very few sectors in mountainous areas where special procedures apply, the MVAs used by controllers already enjoy a buffer beyond any effect caused by the worst-case low temperature day. It is for this reason that the pilot is not required to perform low temperature corrections when receiving vectors from ATC.

This is a cautionary tale. While a temperature-related altimetry error by itself likely won’t result in an accident, the Bell 212 crash reminds us that it can certainly contribute. And special care must be taken in mountainous areas where approach fixes may be high above the airport , knowing that these errors can be very big indeed. And now that it’s a FAA requirement to perform temperature correction at designated airports when conditions demand, it’s the pilot’s legal obligation-responsibility to do so. Remember, temperatures low, look out below.T&T

About the Author

Leave a Reply