July 19, 2013

How “High and Hot” could become “Low and Slow”

"Descend via the SEAVU2 arrival. After SEAVU, cleared for the ILS runway 25L."

It's been a while since I've flown into LAX with any regularity, but after my recent move to the 737, LA seems to be part of just about every trip I fly. On the rare occasion that I flew there as an MD80 FO, I remember getting the same "descend via" clearance that I usually get now...back then ATC rarely let us fly the arrival as it was designed.

In the short amount of time I've been on the 737, I've found SoCal (Southern California) approach to be much better about issuing a “descend via” clearance and allowing the pilot to fly the procedure as it was intended.   Instead of repeated vectors off the arrival, constant speed adjustments and last minute runway changes, the controllers generally allow us to fly the arrival as published. Of course, that isn't always the case, but they definitely don't mess with us as much as they used to.

Descend via?

In case you aren't familiar, a “descend via” clearance tells a pilot to descend using pre-determined parameters as laid out in a published procedure. Below is a copy of the SEAVU arrival and the ILS approach to runway 25L...make note of the crossing restrictions over multiple points on the arrival.  Instead of issuing a separate clearance for each of the restrictions, an air traffic controller could tell a pilot to "descend via the SEAVU" and the pilot would know to make each speed and altitude depicted on the chart.  This simplifies the process for both pilot and controller and reduces congestion on the approach frequency.

This arrival ends at the SEAVU intersection...the approach below begins at the same point.

High and Hot

Yesterday, I approached Los Angeles via the SEAVU TWO arrival and joined the localizer to runway 25L. This is what we normally fly and it's what I had loaded into the FMS before we took off three hours earlier. I was a little outside the HUNDA intersection descending via the arrival with VNAV (Vertical Navigation) engaged and a clearance to intercept the glide-slope when SOCAL instructed me to maintain 4,000 feet.  There was a helicopter crossing the approach corridor, so I had to level off momentarily to avoid the conflict. Even though I was level for less than a minute, it was enough to put me too high and too fast on the approach...and that close to the airport, the options to get down were few.

I should mention that it's standard practice at most airlines to be fully configured for landing, on speed, on glide-slope with stabilized thrust by 1,000 feet above touchdown. We call this a "stabilized approach" and it's very important. Statistically speaking, an airplane is far more likely to execute a successful approach with a touchdown in the landing zone if the airplane meets these parameters.

The short time spent at 4,000 feet seemed like an eternity.  I was re-cleared for the approach a few seconds after the glide-slope indicator slipped out of view, but the damage was done to my carefully planned approach. I was full scale high and although I had started to slow during the level off, I hadn't bled off enough airspeed to facilitate a normal descent.  With that in mind, I reduced thrust to idle, extended the speed brakes, started down and asked the captain to lower the landing gear.

I’ve found the 737 landing gear to be a pilot's best friend when "getting down and slowing down" needs to happen in the approach environment. The gear on many airplanes is loud and causes buffeting at higher airspeeds. This was the case on the MD80, so I rarely used the gear as I did today. On the 737, the gear is hardly noticeable even when lowered at high speed. It's also quite effective at slowing the airplane.

At that point, the flaps were at 10 degrees and the gear was down. I called for flaps 15, 25 then flaps 30 as I slowed below the maximum extension speed for each setting. The glide-slope indicator was still off-scale high, but I thought I had time to get down, so I continued the approach. 

This is where "high and fast" could easily become "low and slow.”

With the engines at idle thrust and the airplane “dirty” (gear, flaps and spoilers extended), the airplane was coming down like a rock.  The airspeed was almost where I wanted it as the glide-slope indicator came off the peg.  I waited until I was about "one dot" high on the the glide-slope then retracted the spoilers and advanced the throttles to spool up the engines.  I met the stabilized approach requirements before reaching 1,000 feet...but just barely.

Jet engines are not like the engine in your car.  When you step on the gas in your car, you get instant power.  When you push the throttle up in a small piston engine airplane, you get a similar response.  But when you request more power from a turbine engine running at idle…be prepared to wait.

Most modern turbine engines have a design feature that helps to mitigate this issue.  The CFM56 engine installed on the Boeing 737 is equipped with Electronic Engine Control (EEC) that automatically sets three minimum idle speeds…ground minimum, flight minimum and approach idle. Ground minimum is selected for ground operations and flight minimum is selected for most phases of flight. Approach idle is selected in flight if flaps are in the landing configuration or engine anti-ice is ON for either engine. At the same airspeed and altitude, engine RPM will be higher for approach idle than for flight minimum idle. This higher RPM improves engine acceleration time in the event of a go-around and would help to protect the pilot in the situation I've described.

A dangerous transition.

It’s the transition from a high rate of descent with engines at idle thrust and decreasing airspeed to a stabilized approach that can be especially dangerous when close to the ground.  If I had been a little higher…if I had ignored the stabilized approach requirements…if I had failed to advance the throttles before really needing the thrust…I could easily have intercepted the glide path well below the 1,000 foot mark and might not have been able to reverse the downward and slowing trend before impacting the ground well before the desired landing zone. 

The reasons for stabilized approach requirements are obvious.  If this event occurred somewhere above 1,000 feet, there would be plenty of room for recovery.  If I had been late to advance the throttles and retract the spoilers, but pulled back on the controls to level off or intercept the glide-slope, the aircraft would immediately slow below the desired approach speed...it's entirely possible that the lag in spool up time on the engines would allow the aircraft to slow dangerously close to stall speed.  Low and Slow.

The airplane is smarter than you might think.  Given room for recovery, the auto throttle system would advance thrust on its own if I neglected to do it myself.  If the auto throttle system was not on or not working properly, the control column would physically shake as I approached a stall.  If I continued to slow, the airplane would automatically push the nose over in an attempt to gain airspeed and depending on the airplane, this would all be accompanied by lights and aural warnings.  With enough altitude, it would be easy to recover from such a condition...but too close to the ground? The same scenario could end in catastrophe.

Obviously, the crash of Asiana 214 was the impetus for this post.  To be clear, while I have opinions on the matter, I do not know, with any certainty, what happened in San Francisco on July 6th and make no claim that what I describe here occurred on that flight.  The NTSB is hard at work and will determine without a shadow of doubt what happened and what needs to be done to make sure it never happens again.  It was the picture below that really got me thinking.


Examples of this type of accident are rare.  Below is a list of documented examples.


The graphic above is a list of "accidents," which are defined differently than "incidents."  Here is how the FAA defines both:

Aircraft accident - an occurrence associated with the operation of an aircraft which takes place between the time any person boards the aircraft with the intention of flight and all such persons have disembarked, and in which any person suffers death or serious injury, or in which the aircraft receives substantial damage.

Incident - an occurrence other than an accident associated with the operation of an aircraft, which affects or could affect the safety of operations.

I point out the difference only because I'm personally aware of a few incidents that were left off the list due to minimal damage to the aircraft and the lack of injuries sustained.  

The moral of the story is not only to make a stabilized approach the goal, but to settle for nothing less.  Regardless of the airplane you fly or the company you work for, safety is the number one goal.  If you’re too high or too low…too fast or too slow…off course or whatever…go around and try it again. 

7 comments:

  1. Brad, thanks for the nice flying story. Pretty recovery to stabilized conditions!

    I recall that Delta L-1011 accident at DFW and, IIRC, it was a classic micro-burst accident -- which leads me to wonder whether the improved micro-burst avoidance technology and procedures that followed can account (in part) for the gap of nearly 20 years prior to Asiana 214.

    Maybe short-landing accidents back in the day tended to result from things like wx encounters, but those guys could fly a visual approach. Today, maybe we've traded causal factors.

    Just a thought (not a happy one).

    Keep safe,

    Frank

    ReplyDelete
  2. Great read Brad!! Thanks for the really technical info! Explains a great deal!!!

    ReplyDelete
  3. Yep, a good post with lots of info - I can in no way be described as anything but an interested armchair observer but my money is on a slam dunk, I reckon they were high but on speed and leveled off on the money but bled off too much airspeed as a result of the level-off manoeuvre before they knew it.....

    Always enjoy your posts!

    Dave from the UK

    ReplyDelete
  4. An excellent post, sir. I too cannot speculate about the July 6 event, but, "If the shoe fits..." Thanks. -C.

    ReplyDelete
  5. Great post Brad. Really like to see pilots with the mentality "better safe than sorry."

    -Swayne

    ReplyDelete
  6. Excellent idea for a post! Obviously we have to bow to the experts at NTSB rather than "armchair speculate," but "If it quacks like a duck..."

    ReplyDelete
  7. Hahaha, it's probably us single-engine VFR pilots that causes you guys to be diverted left and right when coming in to LAX. I would say though that it appears that the amount of air traffic in the Socal tracon area has either decreased or they have gotten more efficient. I remember my first ever solo cross-contry from KRNM to KSNA when I requested VFR flight following, the response I got was "all VFR requests please call back in 10 minutes."

    ReplyDelete