Departure Obstacle Analysis

I was in Aspen about fifteen years ago, sitting in the FBO with at
least ten other crews all doing the same thing: looking at the
overcast. The Obstacle Departure Procedure for the Aspen-Pitkin
County/Sardy Field Airport (KASE) simply says, "use SARDD
DEPARTURE." That departure procedure requires the weather be
at least 400-1 and mandates a climb of at least 460 feet per
nautical mile all the way up to 14,000 feet.
"If we can't see the obstacles," I explained to our lead passenger, "we have to out-climb
them. Our Gulfstream is too heavy to do that so we have to wait until the weather
improves to VFR." Just then all heads in the FBO turned to the runway to see another
Gulfstream take off and disappear into the clouds. Was that crew operating foolishly or
was I being overly cautious? I made it my highest priority to figure this out: what is the
best strategy when dealing with departure obstacles?

Eddie Sez:

If I asked five pilots how to best deal with departure obstacles, I got five different
answers. Even the supposed experts at training centers and commercial vendors
couldn't agree on the basics, much less the solution. The situation isn't much better
today, and no wonder, this is complicated!
Everything here is from the references shown below, with a few comments in orange.

Photo: Aspen Airport at Night, 90-second exposure (Courtesy Tom Cuccio)

• Issue: All-Engines-Operating (AEO) versus One-Engine-Inoperative (OEI) —

Under ICAO and TERPS, obstacle departure procedures assume AEO, but most
of us only have OEI performance numbers.
• Issue: Regulations versus Regulations — There are at least seven FARs, two
ICAO documents, a U.S. Advisory Circular and the five-hundred-plus pages of
that FAA Order known as TERPS all competing for your attention.
• Issue: ICAO versus TERPS — The ICAO and FAA approach the obstacle
avoidance strategy with similar philosophies but some of the terminology is
slightly different and the vertical clearances are very different.
• Issue: The Low, Close-In Problem, Theory versus Reality — Published climb
gradients are not adjusted for low, close-in obstacles, so you must either "see
and avoid" or come up with another way to avoid. This would fine except the
location descriptions are too general to be of any use.
There are many strategies when it comes to dealing with this problem, some better than
others. I'll group them all into three major categories, look at the pros and cons, and
then recommend a strategy.

• The Old School Strategy — Use AFM performance One-Engine-Inoperative

(OEI) performance data to meet published All-Engine-Operating (AEO) climb
gradients. You leave a lot of payload (fuel or passengers) behind, but at least you
are safe. Or are you?
• A Low Tech Strategy — Reduce the Vertical Margins. You can, without any
computer software or even a terrain map, reduce the vertical margins of the ODP
and improve your takeoff performance. Of course this method has its own
complications and does not address the low, close-in obstacle problem.
• A High Tech Strategy — A Terrain Data Base with Obstacle Analysis. Many
major airlines recognized the solution decades ago: simply map the terrain and
all obstacles near airports, determine how accurately aircraft can navigate on
takeoff following an engine failure, and "thread the needle." These days we have
highly accurate terrain maps of most of the earth, derived from satellite imagery.
We have huge computer-driven data bases with digital representations of the
terrain. And, pulling it all together, we have computer software to determine
exactly how much payload aircraft can depart from virtually any airport, and still
avoid all departure obstacles. There does remain, however, one issue.

Recommendations — You can use any of these solutions and probably be okay. Of
course if you hit something you will not be okay and the regulatory agencies will find a
way to blame you. You can improve your odds against the obstacles (and the system),
here is my recommendation.
Gray Area — There is one last source of confusion when operating daily without losing
an engine but always being mindful of losing one when obstacles are a factor. If you
lose an engine at V1 you climb out at V2 until the obstacle is beat. Easy. If you don't lose
an engine you might be tempted to accelerate to 250 knots, cleaning up the flaps as
quickly as you can. You are on two engines and you've made sure you can beat the
obstacles with a normal climb. Or have you? And what happens if you lose that engine
after you brought the flaps up? Now V2 has gone out the window. You need to think
about this. Here's my answer.
This can be a very complex subject because there are so many rules from such a
variety of sources. I've tried to simplify things by quoting only the relevant regulatory
passages but I link to more complete coverage of those passages at the bottom of this
page under Source Extracts.

Issue: All-Engines-Operating (AEO) versus One-Engine-

Inoperative (OEI)
 Photo: American Airlines MD-82 engine fire, September 2007, St. Louis, from the
Associated Press.

Departure procedures assume AEO

• Both U.S. and ICAO departure procedures are built on the assumption that
everything is working, including the engines. The U.S. departure procedure rules
are covered in the United States Standard for Terminal Instrument Procedures
(TERPS), Federal Aviation Administration 8260.3B, Volume 1, ¶201. (Full
extract: TERPS Volume 1, ¶201.)

These criteria are predicated on normal aircraft operations for considering

obstacle clearance requirements.

• The ICAO departure procedures are covered in the ICAO Doc 8168 - Aircraft
Operations - Vol II - Construction of Visual and Instrument Flight Procedures,
Procedures for Air Navigation Services §3, ¶1.7.1. (Full extract: ICAO Doc 8168
Vol II, §3, ¶1.7.1.)

The design of procedures in accordance with this section assumes normal

operations and that all engines are operating.

Airworthiness standards assume OEI

 • Both U.S. and ICAO aircraft certification rules for transport category aircraft
assume failure of the critical engine. The U.S. rules are covered by 14 CFR 25,
§25.111. (Full extract: 14 CFR 25.111 Takeoff Path.)

The airplane must be accelerated on the ground to VEF, at which point the critical
engine must be made inoperative and remain inoperative for the rest of the
takeoff.

• The ICAO Airworthiness rules are covered by ICAO Annex 8 - Airworthiness of

Aircraft, Part IIIA, ¶2.2.3. (Full extract: ICAO Annex 8, Part IIIA, ¶2.2.3.)

The take-off path shall comprise the ground or water run, initial climb and climb-
out, assuming the critical engine to fail suddenly during the take-off.
Some pilots dismiss 14 CFR 25 and ICAO Annex 8 as not really applicable to
flight operations since they have more to do with aircraft certification than pilot
procedures. True or not, these rules cause aircraft manufacturers to focus only
on OEI data so most aircraft certified under these rules present only OEI takeoff
performance data. This greatly impacts turbine aircraft performance planning,
especially when dealing with takeoff obstacles. This is why the classic, "old
school" method of dealing with departure obstacles is to use engine-out data. (In
most aircraft that is all you have.)

Issue: Regulations versus Regulations

 Photo: Aeroflot IL-96-300 departing Salzburg (LOWS), from Hajdufi Gabor (with
photographer's permission).

U.S. Federal Aviation Regulations (Vertical & Lateral)

• If you are operating under U.S. commercial regulations you must use an obstacle
clearance or avoidance procedure that ensures you clear the obstacles by
specific margins. (Full extract: 14 CFR 135, §135.379.)

For an airplane certificated after September 30, 1958 (SR422A, 422B), that
allows a net takeoff flight path that clears all obstacles either by a height of at
least 35 feet vertically, or by at least 200 feet horizontally within the airport
boundaries and by at least 300 feet horizontally after passing the boundaries.

• These aren’t very large clearances — basically 35 feet vertically and 300 feet
horizontally — but you do have to clear the obstacles.
• What about non-commercial operators? You still need to avoid hitting anything
because of the so-called "reckless" rule.

(Full extract: 14 CFR 91, §91.13.)

No person may operate an aircraft in a careless or reckless manner so as to
endanger the life or property of another.
 • Some operators call this the "scrape paint" rule. You have to "just" clear the
obstacle vertically and horizontally. For the sake of the discussion that follows,
we'll assume our 14 CFR 91 operators have the good sense to strive for the
same obstacle clearance standards of their commercial peers.
• It became obvious these margins were far too thin with the dawn of large jet
airliners. At first the rules were tailored for the Boeing 707 but eventually those
rules were tossed out in favor of the so-called "net takeoff flight path" flight path
instituted by 14 CFR 25. While this regulation never mentions a "gross" flight
path, it does have the following to say about the net flight path. (Full extract: 14
CFR 25, §25.115.)

The net takeoff flight path data must be determined so that they represent the
actual takeoff flight paths reduced at each point by a gradient of climb equal to—
(1) 0.8 percent for two-engine airplanes;
(2) 0.9 percent for three-engine airplanes; and
(3) 1.0 percent for four-engine airplanes.

• This is misunderstood by many as "test pilot versus real pilot" rules. That is
completely wrong. The "gross path" is what the airplane actually does with you at
the controls. In fact, you would be better off if you replaced the word "gross" with
the word "actual" when it comes to takeoff climb performance. The regulations
and your flight manual are based on the actual numbers minus the shown
reductions. You can consider this a "safety pad." You and the airplane produce
gross (actual) numbers, the books are based on net numbers.
• Does that mean you can simply add the net reduction back in and call it good?
Or just subtract it from the obstacle departure procedure gradient? Why not just
subtract 0.8 percent from the required climb gradient in a two-engine airplane?
That gains you (0.008)(6076) = 48 feet every nautical mile! But what about
winds? What about an aircraft that might not be as clean as the day it was
certified? What about pilot technique? In my opinion, it is foolish to give up this
margin.

There are those who argue that you don't really need to follow anything in a departure
procedure if you are flying under Part 91. The rules, after all, are for the commercial
guys. The shortest regulation in the book tells us otherwise . . .
[15 CFR 97, §97.1]
(a) This part prescribes standard instrument approach procedures to civil airports in the
United States and the weather minimums that apply to landings under IFR at those
airports.
(b) This part also prescribes obstacle departure procedures (ODPs) for certain civil
airports in the United States and the weather minimums that apply to takeoffs under IFR
at civil airports in the United States.
In the United States, every instrument approach, arrival, and departure procedure is
regulatory. If you are flying under instrument flight rules, you have to obey everything on
that plate.

U.S. TERPS (Vertical)

If you are flying an instrument procedure in the United States, chances are it was
designed under the criteria set down by the United States Standard for Terminal
Instrument Procedures, also known as "TERPS" or Federal Aviation
Administration 8260.3B. TERPS has other names with various branches of the
U.S. military and it is also used by various foreign governments without their
own airspace design authorities. If you are flying an obstacle departue procedure
designed under TERPS, there are a few things you need to know about
departure obstacle avoidance . . .

Illustration: Climb Segment, from TERPS, ¶202b, figure 1-3.

• When flying a TERPS-designed ODP, you are given a minimum Climb Gradient
(CG) of 200 feet per nautical mile and a minimum Required Obstacle Clearance
(ROC) of 48 feet per nautical mile. These values of CG and ROC remain so long
as no obstacles (with an exception to be covered later) penetrate an Obstacle
Clearance Surface (OCS) derived as 152 feet per nautical mile. If an obstacle
does penetrate the OCS (other than the exception to be covered shortly), the CG
and ROC are increased to maintain at least a 24 percent buffer between the flight
path and the obstacle. These rules in TERPS are covered in Volume 1, ¶203.
(Full extract: TERPS, Volume 1, ¶203.)

For TERPS purposes, the MINIMUM climb gradient that will provide adequate
ROC in the climb segment is 200 ft/NM.
 The vertical distance between the climbing flight path and the OCS is ROC. ROC
for a climbing segment is defined as ROC = 0.24 CG . This concept is often
called the 24 percent rule.
Where an obstruction penetrates the OCS, a nonstandard climb gradient (greater
than 200 ft/NM) is required to provide adequate ROC.
The nonstandard ROC expressed in ft/NM can be calculated using the formula:
(0.24 h) ÷ (0.76d) where "h" is the height of the obstacle above the altitude from
which the climb is initiated, and "d" is the distance in NM from the initiation of
climb to the obstacle.

• Note that the ROC is defined as being 24 percent of the CG. The minimum ROC
comes to 48 feet at 1 nautical mile because (0.24)(200) = 48 feet. If the CG
increases, the ROC increases too. So, for example, if you have a 300 feet per
nautical mile climb gradient, your ROC at 1 nautical mile will be (0.24)(300) = 72
feet.

U.S. TERPS (Lateral)

Illustration: Initial Climb Area, from TERPS, Vol 4, figure 1-5.

 • While most pilot texts covering departure obstacles pay heed to the need to
maintain an increasing flight path vertically over obstacles, many fail to
consider which obstacles. If your takeoff flight path takes you right over an
obstacle that is obvious enough. But what if that obstacle is a mile to the left once
you've already traveled five miles? Do you need to clear that obstacle? It
depends on the rules that apply. Under TERPS, the Initial Climb Area (ICA) starts
out at 500 feet left and right of runway centerline at the Departure End of Runway
(DER). It normally ends 2 nautical miles later but could be extended for 10
nautical miles. You will have to consider all obstacles up to 500 feet of centerline
at first, extending to a width that can vary from 3,756 feet to just under 3 nautical
miles. All of this is covered by TERPS, Volume 4, ¶1.6.

(Full extract: TERPS, Volume 4, ¶1.6.)

The Initial Climb Area Baseline (ICAB) is a line extending perpendicular to the
runway centerline 􏱉 500 at DER. The splay of 15° and length of the ICA
determine its width. The ICA length is normally 2 NM. The ICA may be extended
beyond 2 NM 􏱉to maximum length of 10 NM.

• This is the simplest and narrowest of the possible departure procedures; the
lateral margins increase with turns and can vary with available course guidance.
Having to out-climb an obstacle that is over two miles away laterally would seem
nonsensical. U.S. rules recognized this with the adoption of Advisory Circular
120-91.

U.S. AC 120-91 (Lateral)

• It has been rather obvious to the airlines that U.S. TERPS departure procedure
designs are overly restrictive, but the narrow margins afforded by commercial
regulations are too risky. Over a series of many years of collaborative research
the FAA released Advisory Circular 120-91, Airport Obstacle Analysis, which
specifies a more realistic standard when it comes to lateral margins.

(Full extract: AC 120-91, ¶ 10, 11, 12.)

The Area Analysis Method defines an obstacle accountability area (OAA) within
which all obstacles must be cleared vertically. The OAA is centered on the
intended flight track and is acceptable for use without accounting for factors that
may affect the actual flight track relative to the intended track, such as wind and
available course guidance.
During straight-out departures or when the intended track or airplane heading is
within 15 degrees of the extended runway centerline heading, the minimum width
of the OAA is 200 feet on each side of the intended track within the airport
boundaries, and 300 feet on each side of the intended track outside the airport
boundaries. During departures involving turns of the intended track or when the
airplane heading is more than 15 degrees from the extended runway centerline
 heading, the maximum width of the OAA is 3,000 feet on each side of the
intended track.

• If you operate using the guidance of AC 120-91, you can narrow your lateral
margins from nearly 3 miles down to only 3,000 feet of each wing tip. In other
words, you no longer have to restrict your takeoff weight to out-climb that
obstacle sitting over a half-mile to one side.

ICAO Regulations (Vertical)

• The ICAO also provides for a net takeoff flight path, similar to 14 CFR 25, that
requires the AFM performance be reduced by set margins. If a factor is applied
during certification, such as is the case for transport category aircraft certified in
the United States, that factor is used. If no factor is given, the ICAO provides a
net factor for 2 and 4 engine aircraft that are more conservative. If your aircraft
was certified under 14 CFR 25, you will use the FAA rules.

(Full extract: ICAO Annex 6, Part I, ATT C.)

The term “net take-off flight path”, as relating to the aeroplane, has its meaning
defined in the airworthiness requirements under which the aeroplane was
certificated. If this definition is found inadequate, then a definition specified by the
State of the Operator should be used.
The net take-off flight path is the one-engine-inoperative flight path which starts
at a height of 10.7 m (35 ft) at the end of the take-off distance required and
extends to a height of at least 450 m (1,500 ft) calculated in accordance with the
conditions of 2.9, the expected gradient of climb being diminished at each point
by a gradient equal to: 0.5 per cent, for aeroplanes with two engines, 0.8 per
cent, for aeroplanes with four engines.

• As with the 14 CFR 25 definition, aircraft manufacturers are compelled to publish

AFM performance numbers with the net flight path performance numbers. What
this means to the pilot is that the airplane will actually climb better than predicted
by their manuals. Some pilots talk of increasing the AFM performance by this so-
called "net takeoff flight path" factor. I recommend you do not do this. The added
margin becomes a safety factor to account for changes in wind and temperature,
as well as variation in pilot technique. Besides, you have other vertical margins to
consider.
 Illustration: Procedure Design Gradient, from ICAO Doc 8168 Vol II, figure I-3-2-
2.

• If you are flying an obstacle departure procedure designed under ICAO rules,
you are given a 0.8 percent margin between the Procedure Design Gradient
(PDG) and the obstacle. This margin is the Minimum Obstacle Clearance (MOC),
akin to the TERPS Required Obstacle Clearance (ROC) but smaller.

(Full extract: ICAO Doc 8168 Vol II, ¶2.)

The standard procedure design gradient (PDG) is 3.3 per cent. The PDG begins
at a point 5 m (16 ft) above the departure end of the runway (DER).
The standard PDG provides an additional clearance of 0.8 per cent of the
distance flown from the DER, above an obstacle identification surface (OIS). The
OIS has a gradient of 2.5 per cent.
Where an obstacle penetrates the OIS, a steeper PDG may be promulgated to
provide obstacle clearance of 0.8 per cent of the distance flown from the DER.
The minimum obstacle clearance (MOC) in the primary area is 0.8 per cent of the
distance flown from the DER. The MOC is zero at the DER.

• ICAO uses a 3.3 percent line drawn from 5 meters above the departure end of
the runway to construct what they call the Procedure Design Gradient (PDG).
This is identical to the U.S. TERPS 200 feet per nautical mile climb gradient,
since 100 (200 / 6076) = 3.3 percent, except for the matter of how far above the
runway this gradient begins. The Obstacle Identification Surface (OIS) is used to
 identify if adjustments are needed to the standard PDG. The OIS has a gradient
of 2.5 percent, which is identical to the U.S. TERPS 152 feet per nautical mile
Obstacle Clearance Surface (OCS), since 100 (152 / 6076) = 2.5 percent
• The ICAO adjusts the PDG upwards to maintain a 0.8 percent Minimum Obstacle
Clearance (MOC) above all obstacles. Unlike the rule in U.S. TERPS, the ICAO
will allow the climb gradient to change back to 3.3 percent once the obstacle has
been passed.

ICAO Regulations (Lateral)

Illustration: Straight departure area without track guidance, from ICAO Doc 8168 Vol II,
figure I-3-3-1.

• ICAO obstacle departure procedures are built with lateral margins that are even
more generous than those found in U.S. TERPS.

(Full extract: ICAO Doc 8168 Vol II, ¶3.2.4.1.)

The area begins at the DER and has an initial width of 300 m (Cat H, 90 m).
The departure procedure ends at the point where the PDG reaches the minimum
altitude/height authorized for the next phase of flight.
 • This is the simplest and narrowest of the possible departure procedures; the
lateral margins increase with turns. Note that it begins at 150 meters (492 feet)
either side of centerline and gets as wide as 150 + (20,000) tan(15) = 5,659
meters (just over 3 nautical miles).
• The ICAO narrows the area an operator must consider when flying a turbine-
powered aircraft over 5,700 kg maximum certificated takeoff mass.

(Full extract: ICAO Annex 6, Part I, ATT C.)

No aeroplane should commence a take-off at a mass in excess of that shown in
the flight manual to correspond with a net take-off flight path which clears all
obstacles either by at least a height of 10.7 m (35 ft) vertically or at least 90 m
(300 ft) plus 0.125D laterally, where D is the horizontal distance the aeroplane
has travelled from the end of take-off distance available.
Where the intended track does not include any change of heading greater than
15°, for operations conducted in VMC by day, or for operations conducted with
navigation aids such that the pilot can maintain the aeroplane on the intended
track with the same precision as for operations specified in 5.1.1 a), obstacles at
a distance greater than 300 m (1,000 ft) on either side of the intended track need
not be cleared.
Where the intended track does not include any change of heading greater than
15° for operations conducted in IMC, or in VMC by night, except as provided in
5.1.1 b); and where the intended track includes changes of heading greater than
15° for operations conducted in VMC by day, obstacles at a distance greater than
600 m (2,000 ft) on either side of the intended track need not be cleared.
Where the intended track includes changes of heading greater than 15° for
operations conducted in IMC, or in VMC by night, obstacles at a distance greater
than 900 m (3,000 ft) on either side of the intended track need not be cleared.

• For turbine powered aircraft that weigh more than 5,700 kg (12,566 pounds) the
margins are narrowed considerably, depending on turns and course guidance.
The maximum lateral width is 3,000 feet but can be as little as 1,000 feet.

Issue: ICAO versus TERPS

Vertical
 Figure: Climb gradient requirement comparison, from Eddie's notes.
There are a few differences in the way the ICAO and the U.S. FAA (through TERPS)
address obstacles that will impact what you can do to avoid them. You cannot address
the vertical component of obstacle avoidance without also considering the lateral
component. If the obstacle is right below you, of course, you must out-climb it. What if it
is 100 feet to your right? How about a mile? For now let's table the lateral discussion
and simply look at the vertical.
The minimum climb gradient. Both ICAO and TERPS specify a minimum climb
gradient for all departure procedures. The ICAO calls this the Minimum Procedure
Design Gradient (PDG) and says it can never be less than 3.3 percent. In the U.S.,
TERPS calls this the Minimum Climb Gradient (CG) and says it can never be less than
200 feet per nautical mile. These values are about the same, since 200/6076 = 0.033
and that is another way of writing 3.3 percent.
The obstacle clearance / identification surface. Both ICAO and TERPS specify a
surface below the aircraft's path that identifies a zone where obstacles cannot penetrate
without having to change the climb gradient. There is an exception to this for Low,
Close-In Obstacles, but more on that later. The ICAO calls this the Obstacle
Identification Surface (OIS) and defines it as a surface starting at the Departure End of
Runway (DER) inclined upward by 2.5 percent. In the U.S., TERPS calls this the
Obstacle Clearance Surface (OCS) and defines it as a surface that starts at the DER
inclined upward by 152 feet per nautical mile. The values are about the same, since
152/6076 = 0.025 and that is another way of writing 2.5 percent.
Minimum / Required Obstacle Clearance. If you take the minimum climb gradient and
subtract the obstacle surface you get the safety margin between the two. If you simply
look at the minimum climb gradient where no obstacles penetrate the identification /
clearance surface, both ICAO and TERPS provide for about the same margin. Under
ICAO you have 3.3 - 2.5 = 0.8 percent. Under TERPS you have 200 - 152 = 48 feet per
nautical mile, and that comes to (48 / 6076) = 0.0079, or about 0.8 percent. But as the
climb gradient increases, the TERPS value increases. Here TERPS jumps from using
feet per nautical mile to percentages, but the TERPS percentage is different.

• Under ICAO, a Minimum Obstacle Clearance (MOC) of 0.8 percent of the

distance from the DER is specified. Mathematically, MOC = (0.008) (d).
• Under TERPS a Required Obstacle Clearance (ROC) of 24 percent of the Climb
Gradient is specified. Mathematically, ROC = (0.24) CG.

Adjusting climb gradient for obstacles. If an obstacle, other than a Low, Close-In
Obstacle (more on that later) penetrates the OIS / OCS, the procedure's climb gradient
must be raised to preserve the MOC / ROC.

• Under ICAO, the 0.8 percent MOC is added to the gradient created by the
obstacle. If, for example, a line from the DER to the obstacle is 5 percent, the
Procedure Design Gradient is raised to 5.8 percent.
• Under TERPS, the Climb Gradient is adjusted to the following formula:

Climb Gradient=h (feet)0.76d (nautical miles)Climb

Gradient=h (feet)0.76d (nautical miles)
You can compute the resulting ROC = (0.24) (CG), which is why they call this the
"24 percent rule." Yes, the ROC is larger than the MOC, but the numbers 24 and
0.8 exaggerate the difference because they are percentages of different things.

Vertical Example
Let's say we have an obstacle that is 1500 feet above and 5 nm (30,380 feet) away from
the DER.
Under ICAO, the obstacle has a gradient of (1500 / 30380) = 0.0494, or 4.94 percent.
The MOC is always 0.8 percent so our PDG is 4.94 + 0.8 = 5.74 percent. Our height
above the obstacle would be (0.0574)(30380) - 1500 = 244 feet.
Under TERPS, the climb gradient is h / (0.76 d), or 1500 / (0.76 x 5) = 395 feet per
nautical mile. (That's 6.5 percent, much higher than the ICAO PDG.) So our ROC =
(0.24) (395) = 95 feet per nautical mile. At 5 nm, our height above the obstacle will be
(5)(95) = 475 feet. You can also derive this by figuring your altitude (5)(395) = 1,975,
subtracting the obstacle height (1,500 feet) to arrive at the same answer, 475 feet.
Is 244 feet (ICAO) or 475 feet (TERPS) a comfortable margin? Keep in mind that if your
multi-engine turbine aircraft was certified under 14 CFR 25, you will also have the net
takeoff flight path margin. A twin-engine aircraft, for example, will be (0.008) (5) (6076) =
243 feet higher than the AFM states, unless there is a tailwind or temperature inversion.
Only you can decide this, but you will need to give it some thought. Chipping away at
this vertical margin is a fundamental step in the techniques to follow.

Lateral

Figure: Obstacle area lateral comparison, from Eddie's notes.

Obstacle departure procedures are designed with very wide lateral tolerances under
both ICAO and TERPS, which means those minimum climb gradients could be
unnecessarily high because they are considering obstacles miles away from course
centerline. Of course this was necessary back when an aircraft climbing into a cloud
deck was lucky to be within a mile of course. But these days? If you have an airplane
with an instantaneous readout of "position uncertainty" you very seldom see your
airplane more than 0.05 nautical miles off course. Three hundred feet! While departure
procedures continue to be built off these wide lateral areas, we as pilots are allowed to
narrow our gaze if we have a plan to do that.
Original ODP Designs. TERPS procedure construction can be very complicated; the
lateral margins vary with distance from the departure end of the runway, relationship to
the airport boundary, any turns, and available track guidance. The lateral margin starts
at 200 feet either side of runway centerline and quickly expands by thousands of feet, to
as much as 3 miles. ICAO procedure construction mimics TERPS in many ways and
becomes almost as wide. Unless the procedure says otherwise, the climb gradient on
these procedures could be based on obstacles that you will have no chance of seeing.
Tighter Lateral Margins. ICAO Annex 6 narrows the lateral margin for large (more than
5,700 kg, about 12,500 lbs.) turbine aircraft. The margin can be as tight as 1,000 feet
but will be no more than 3,000 feet (about a half nautical mile), depending on course
guidance, turns, and distance from the runway. U.S. Advisory Circular 120-91 provides
a method of applying a obstacle clearance area that is much more narrow than TERPS
and is almost as narrow as the tightest ICAO margin. If an aircraft can maintain course
within 3,000 feet (about a half nautical mile), the result can be a significant increase in
payload.

Lateral Example

Figure: 3-D Terrain Model, TERPS, Aft View, from Eddie's notes.
Let's say you are departing in a two-engine aircraft from an airport that leads into a
valley with what looks to be a challenging obstacle departure procedure. The SID says
you need to climb at 400 ft/nm to an altitude that is 4,000 ft above the departure end of
the runway. Looking at the chart you see a number of mountains and it appears the
greatest problem will be around 10 nm after takeoff about 3 nm to the right. The
departure takes you right down the middle of the valley, so can you improve your
situation by keeping to the course centerline better than 3 nautical miles? How high
above the obstacle will you really be?

• The formula for finding the climb gradient in a TERPS procedure is CG = h /

(0.76 x d), where h is the obstacle height in feet and d is the distance from the
 DER in nautical miles. We can run the formula backwards to find h = CG (0.76) d
= 400 (0.76) (10) = 3,040 feet above the DER.
• The procedure required obstacle clearance is ROC = (0.24) CG = (0.24) (400) =
96 ft/nm, which means at 10 nm it will be 960 feet.
• Since the aircraft was certified under 14 CFR 25, you also know it will outperform
the AFM net takeoff flight path by 0.8 percent, which means at 10 nm it will be
(0.008)(10)(6076) = 486 feet higher than the charts say.
• While TERPS assumes the departure begins at DER on the runway, your aircraft
manuals are usually predicated on 35 feet. If the aircraft was certified with wet
runway takeoff performance, it may be predicated on crossing the DER at 15 feet
in wet runway conditions. Let's say the runway is dry and you cross the DER at
35 feet, which means you will be that much higher abeam the obstacle.
• That means you will cross abeam the obstacle 960 + 486 + 35 = 1,481' higher
than the obstacle.

Combining the Vertical and Lateral Examples

Figure: 3-D Terrain Model, AC 120-91, Aft View, from Eddie's notes.
In our example, if you lose an engine at V1 and manage to keep the aircraft on the
departure procedure's course centerline, you will be 1,481 feet above (vertically) and 3
nautical miles away (laterally). If you don't lose an engine, you will of course be much
higher.
You have several vertical margins at work: the 35 feet required by 14 CFR 135.379
(which you cannot give up), the net takeoff flight path margin (0.8 percent for a two-
engine aircraft), and the 24 percent Required Obstacle Clearance afforded by TERPS.
Giving up the ROC would still leave you a margin of 35 + (10)(0.008)(6076) = 521 feet
above and abeam the obstacle.
We will shortly examine a method that addresses the vertical margins and another that
addresses the vertical and lateral margins. But first, there is a more immediate problem
with all obstacle departure procedures, regardless of the number of engines operating .
..

Issue: The Low, Close-In Problem, Theory versus Reality

Photo: Burbank Runway 33, from Kurt Preisler.

There is a catch when it comes to low, close-in obstacles under both ICAO and U.S.
TERPS departure procedures. You are expected to avoid them using "see and avoid" or
procedural techniques, but you aren't given precise location information. This
information is detailed in Volume 4 of TERPS.
(Full extract: TERPS Vol 4, ¶ 1.4.6)
Where low, close-in obstacles result in a climb gradient to an altitude 200 feet or less
above DER elevation . . . publish a note identifying the obstacle(s) type, location relative
to DER, AGL height, and MSL elevation.
The 200 feet or less implies any obstacles below 200 feet above DER within 1 nm of the
DER because the standard climb gradient is 200 ft/nm. Higher and farther obstacles
require a change to published climb gradients and possibly to takeoff minimums. But
those low, close-in obstacles require only a note.
(Full extract: TERPS Vol 4, ¶ 1.3.1)
Do not publish a CG to a height of 200 feet or less above the DER elevation. Annotate
the location and height of any obstacles that cause such climb gradients.
The same "catch" exists under ICAO.
(Full extract: ICAO Doc 8168 Vol II, ¶2.)
An increased gradient that is required to a height of 60 m (200 ft) or less, (normally due
to low, close-in obstacles) shall not be promulgated. The position and elevation/height
of close-in obstacles penetrating the OIS shall be promulgated
Where it all falls apart with this low, close-in obstacle methodology is the "publish a note
identifying the obstacle(s)" proviso. To see this in action, consider Runway 33 at Bob
Hope Airport (KBUR) at Burbank, California.

Figure: Burbank Airport Takeoff Minimums, (Obstacle) Departure Procedures.

At first reading the published notes seem ridiculous. Can there possibly be a 100'
obstacle 33' from the DER just 30' right of centerline? I would have noticed! The word
"beginning" explains the rationale behind the poorly worded sentence but does little to
help the pilot. Where are these obstacles, exactly? They are not drawn precisely (if at
all) on the airport diagrams and a sectional is useless at this level of detail. The FAA
does offer a digital obstacle file for the United States
at http://www.faa.gov/air_traffic/flight_info/aeronav/digital_products/dof/ but these are
very large, cumbersome, and would take a good computer to really digest. If you
wanted to try, you will see that the file covering KBUR is 790 pages long and includes
these two gems:
06-030661 O US CA BURBANK 34 12 56.17N 118 21 50.28W POLE 1,00050 00846 R
2 C U A 2013106
06-001786 O US CA BURBANK 34 12 52.00N 118 21 41.00W POLE 1,00048 00831 L
1 A U C 2014152
So if you were able to find these two obstacles out of the thousands given, and if you
plotted them, you would see where exactly your low, close-in obstacles are:

Figure: Burbank Airport Low, Close-In Obstacles, from Eddie's notes.

Case solved, right? Not so fast. These were just the two obstacles I found after combing
through the digital obstacle file for several hours. Even if I managed to plot all of them,
the file only includes man-made objects. Can you really be expected to see and avoid
these low, close-in obstacles even if the location data was more precise?
Keep in mind this problem exists even if you load your aircraft to meet TERPS and
ICAO climb gradients because those climb gradients do not consider low, close-in
obstacles.
Is this a problem? Almost never. Almost. In our example, let's say it is raining, the
weather is above standard and you are permitted to leave with the minimum climb
gradient of 300 ft/nm to 5,000 feet. That comes to 300 / 6076 = 4.94 percent. If the
maximum weight for this climb gradient requires a ground run following an engine failure
that equals the runway available, you can expect to cross the DER at 15 feet. A 4.94
percent gradient across a distance of 812 feet results in a climb of (0.0494) (812) = 40
feet. If you cross the DER at 15 feet that means you are at 40 + 15 = 55 feet when you
cross the pole marked as DOF 06-001786, which is 53 feet above the DER. You have a
clearance of 2 feet!

How to out-climb all low, close-in obstacles

Figure: How to out-climb low, close-in obstacles, from Eddie's notes.

If you are departing an airport with ambiguous notes about low, close-in obstacles the
only thing you can be sure of is that they are out there. You can examine the area with
your own eyes, comb through digital obstacle files, or simply out-climb them. If you
cross the DER at 200 feet you know you will be above them all. To do this, you will need
to know how much runway you must have left over when you lift off. You can figure this
out by multiplying 6076 by 200, and dividing all that by the climb gradient in feet per
nautical miles. If, for example, the AFM says you will be climbing at 400 feet per nautical
mile, you will need to have (6076)(200) / (400) = 3038 feet of the runway remaining
when you lift off.
Of course this is a ridiculous solution to what should be a simple problem. If you are
going to be using all of that runway and there is any doubt about the low, close-in
obstacles, the airport manager should be able to help with a more specific idea of where
these obstacles are. There is, fortunately, an even easier solution. (See below, A High
Tech Strategy.)
Now that we've covered most of the issues involved with departure obstacle avoidance,
let's look at three possible strategies.

The Old School Strategy: Use AFM OEI Performance With

Published AEO Departure Procedures
Not too many years ago the vast majority of pilots would tell you that the only way to
legally and safely depart a mountainous airport was to look at the departure procedure,
look at your Airplane Flight Manual (AFM) performance charts, and make everything
agree. So let's try the Aspen Obstacle Departure Procedure (ODP) with my favorite
example airplane, the Gulfstream G450.

Figure: G450 Net Gradient Takeoff Second Segment Chart, KASE, 20 Flaps, 20°C,
from Eddie's notes.
Chasing through the charts it looks like an impossible task. We find the appropriate
charts in G450 AFM, §05-06-00, Figure 3. Our first task is to compute the climb
gradient. The departure tells us we need 460 feet per nautical mile. We know a nautical
mile is 6,076 feet, therefore:

Obstacle Departure Climb Gradient=460 feet / nautical mile6076 feet /

nautical mile=7.6%Obstacle Departure Climb Gradient=460 feet / nautical
mile6076 feet / nautical mile=7.6%
But the Gulfstream manual tells us this is a "gross climb gradient" and that we should
subtract 0.8 percent so that it becomes a net takeoff flight path climb gradient. It seems
the manual wants us to give up this safety margin when flying an obstacle departure
procedure with no further explanation. Since we know there are other safety margins
out there, we dutifully enter the chart with a value of:

"Net" (or Reference) Climb Gradient=7.6−0.8=6.8%"Net" (or Reference)

Climb Gradient=7.6-0.8=6.8%
Chasing up the chart we find we need a "Gross Level-off Height" which the AFM says
we can find by subtract the elevation of the Departure End of Runway (DER) from the
MSL level off:

Gross Level-off Height=14000−7680=6320ftGross Level-off Height=14000-

7680=6320ft
We come up with a transfer scale number of 8.4. Entering the next chart at 20°C and
7600' pressure altitude we find our maximum grossweight will be around 48,000 lbs.
Since our airplane has a BOW of around 43,000 lbs, we won't have enough gas to
make it to Denver with any kind of reserve. Since we are trying to make it back to the
east coast, that only leaves us with one option: wait for the weather to improve so we no
longer have to comply with the climb gradient.

Photo: G450 MCDU Takeoff Data Page 1/3, KASE SARDD Obstacle Limited Page,
from Eddie's aircraft.
Meanwhile your copilot remembers that the aircraft FMS has a performance computer
that can automate those silly spaghetti charts and enters all the appropriate data. With
this FMS you enter climb gradient in feet per nautical mile to a stated elevation (MSL) in
feet.
 Photo: G450 MCDU Takeoff Data Page 1/3, KASE SARDD Obstacle Limited Page,
from Eddie's aircraft.
The answer, unfortunately, comes up very close to those spaghetti charts.
Remember that this number assumes you need to avoid all obstacles within very large
vertical and lateral margins. Though the Gulfstream solution immediately gives up
the net takeoff flight path 0.8 percent performance margin, you still have the 24 percent
Required Obstacle Clearance that isn't really required at all if you lose an engine. There
is room for improvement . . .

A Low Tech Strategy: Reduce the Vertical Margins

The Strategy
There is a strategy in use by many operators and provided for by several commercial
vendors to simply subtract either the ICAO 0.8 percent MOC or the TERPS 24 percent
ROC from the Obstacle Departure Procedure (ODP) climb gradient. Since very few
aircraft manufacturers supply All-Engine-Operative (AEO) takeoff climb performance
data, we've normally ensured we meet the ODP climb gradient with One-Engine-
Inoperative (OEI). We know that if we can make the climb gradient with an engine
failed, we'll have no problems with all engines operating:
 Figure: Meeting the ODP climb gradient with OEI, from Eddie's notes.
You can do this, but there are three caveats:

1. If you lose an engine, you must clear all obstacles by 35 feet vertically and 200, 300,
or up to 3,000 feet laterally depending on your distance from the airport and which
rules apply. (If you've reduced your climb gradient by the MOC or the ROC, you
should be okay since they are designed with this in mind and the net takeoff flight
path margin is still there.)
2. If you lose an engine, you should declare an emergency so ATC knows you will not
be climbing as expected and that they should "clear the way."
3. If you do not lose an engine, you must still make the ODP climb gradient.

So let's say you've figure a new, higher gross weight by reducing the climb gradient by
24 percent on a TERPS departure procedure. Instead of a 400 feet per nautical mile
gradient, for example, you enter your performance computer with (1 - 0.24) 400 = 304
feet per nautical mile. If you lose an engine, you will clear the obstacles and don't have
to worry about the procedure's climb gradient. But if you don't lose the engine, will you
still make the required climb gradient? You are in uncharted territory:
 Figure: Meeting the ODP obstacle clearance gradients with OEI, from Eddie's notes.

Ensuring All-Engine-Operating (AEO) Performance at Higher Weights

If your AFM does not include AEO takeoff climb data you might have a problem, but you
might find something that gives you something that is a conservative analog. In the
G450, for example, we have a chart that produces all engine climb data with the gear
down and flaps fully extended to 39°. If the airplane can make the climb gradient in this
configuration, it can surely do so with the gear up and a smaller angle of flaps:
Figure: G450 All Engine Climb, from G450 AOM, §13-02-10, figure 2.
Even if you don't have any conservative chart that gives you the necessary
reassurance, you might be okay. Let's say, as with the Aspen example, you have a
ODP climb gradient of 7.6 percent and elect to reduce that by the TERPS 24 percent
ROC, lowering your OEI climb gradient to (1 - 0.24) (7.6) = 5.8 percent. You know you
will clear the obstacles because the climb gradient minus the ROC based on that. Now
if you don't lose an engine what is your AEO climb gradient? What follows is my
personal theory.

• If you are flying a two-engine aircraft you are getting half your climb gradient from
each engine. If you lose an engine, your climb gradient decreases by at least 50
percent because you will also have the parasite drag from the wind milling or
seized engine.
• It follows, then, that your all-engine climb gradient will be at least double your
one-engine climb gradient.
• Since you've reduced your target climb gradient by a maximum of 24 percent and
will have double the climb gradient available, you should be okay.
• Since the loss of an engine in a three-engine aircraft results in 33 percent thrust
loss and in a four-engine aircraft results in a 25 percent thrust loss, each
aircraft should be okay since the maximum gradient reduction is 24 percent.

Should? There might be something I haven't thought of here. I've tested this in the
simulator in a GIV, GV, G450, and CL-604. See the G450 results here: Departure
Obstacle Avoidance / Takeoff Climb Performance AEO Versus OEI. I encourage you to
do the same. Our results have been very good. If, for example, the OEI climb gradient
was 8 percent our AEO gradient was easily 20 percent. To do this, have the simulator
operator record the aircraft's flight track and depart on an ODP twice, once with an
engine failed at V1 and once with an all-engine aircraft.

Back to the Aspen Example

We can examine the SARDD obstacle departure procedure from Aspen to better
understand these margins. Recall that we were unable to depart Runway 33 at Aspen
under Instrument Flight Rules because we could not meet the 460 feet per nautical mile
climb gradient all the way to 14,000 feet.
Using terrain mapping software, such as Google Earth, you can draw the obstacle
departure procedure course line from the Departure End of Runway (DER) all the way
to the completion of the procedure. You can also diagram the borders of the obstacle
clearance area and discover the most challenging obstacles are about a mile right of
course. This is a laborious process and no pilot should be expected to do this. But this
examination will help illustrate what exactly is going on when you choose to reduce your
vertical margins to make an obstacle departure procedure climb gradient.

• The published climb gradient is 460 feet per nautical mile, which comes to (460 /
6076) = 7.57 percent
• The Departure End of Runway (DER) is 7,680 feet. We will reach 14,000 feet in
(14000 - 7680) / 460 = 13.74 nm
• A theoretical controlling obstacle height can be derived from the TERPS formula:

Obstacle Departure Climb Gradient=h (height of obstacle above

DER in feet)0.76d (distance from DER of obstacle in nautical
miles)Obstacle Departure Climb Gradient=h (height of obstacle above DER in
feet)0.76d (distance from DER of obstacle in nautical miles)
From this we see h = (0.76) (d) (CG) = (0.76) (13.74) (460) = 4804 feet

• From this we derive the obstacle gradient that is controlling our procedure's climb
gradient, it is:

Theoretical Obstacle
Gradient=4804(13.74)(6076)=5.75 %Theoretical Obstacle
Gradient=4804(13.74)(6076)=5.75 %

• Using Google Earth, we can produce a terrain elevation profile for an on course
departure (shown below in blue) and for one deviates to the right inside the
TERPS obstacle clearance area until it is 1 nautical mile to the right (shown in
red). Right of course we see an obstacle at 9,250' MSL, 4.5 nm from the DER.
This obstacle will be 9250 - 7680 = 1,570 feet above the DER. We can compute
its gradient:

Example Obstacle (1 nm right of course)

Gradient=1570(4.5)(6076)=5.74 %Example Obstacle (1 nm right of
course) Gradient=1570(4.5)(6076)=5.74 %
(Pretty close to the theoretical gradient.)

• We can repeat this process for what appears to be the most challenging obstacle
if the airplane were to remain precisely on course, a peak of 8,700 feet found 7.2
nm from DER. The peak is 8700 - 7680 = 1,020 feet above DER. The gradient of
this obstacle is:

Example Obstacle (on course)

Gradient=1020(7.2)(6076)=2.33 %Example Obstacle (on course)
Gradient=1020(7.2)(6076)=2.33 %
This is less than the 2.5 percent ICAO OIS and the TERPS 152 ft/nm OCS, since
(0.0233)(6076) = 142 ft/nm. If you could remain on course you would only need
the minimum 200 ft/nm (TERPS) 3.3 percent (ICAO) climb gradient!
 Figure: Aspen SARDD obstacle departure gradients, from Eddie's notes.

Reducing Climb Gradient Strategy by Required Obstacle Clearance

(ROC) or Minimum Obstacle Clearance (MOC)
Since this departure procedure was designed using TERPS, one could argue that the
24 percent ROC can be removed in the event of an engine failure and the airplane
would still have the net takeoff flight path safety pad.

• Loading your aircraft to the OEI numbers and the AEO climb gradient, your
altitude over the 4.5 nm obstacle will be (4.5) (460) + 7680 = 9,750 feet.
• If you elect to load your aircraft so as to achieve 24 percent less climb, (1 - 0.24)
(460) = 350 ft/nm, your altitude over the obstacle will be (4.5) (350) + 7680 =
9,255.

Just 5 feet vertical clearance! So will you cross the obstacle right at that altitude? No,
remember your net takeoff flight path factor. A two-engine aircraft will actually be (.008)
(4.5) (6076) = 219 feet above the obstacle. This is precisely the solution favored by
some commercial vendors.
Another option would be to apply the smaller ICAO 0.8 percent MOC to the TERPS
procedure. In our example, the aircraft would be loaded to provide for a climb gradient
of 7.57 - 0.8 = 6.77 percent. You will cross the obstacle at (.0677) (4.5) (6076) + 7680 =
9,531 feet, 281 feet above the obstacle. Once you thrown in the net takeoff flight
path factor, you clear the obstacle by 281 + 219 = 500 feet.

Commercial Options
You can easily compute your own reduced climb gradient by going through the charts
and simply starting with a climb gradient reduced by the 24 percent ROC on a TERPS
procedure or the 0.8 percent MOC on an ICAO procedure. You just need to be careful
how you do that:

• On a TERPS procedure you multiply the published gradient by 0.76, making it 24

percent less. Our Aspen procedure goes from 460 ft/nm to (0.76) 460 = 350
ft/nm. (This equates to 350 / 6076 = 5.8%)
• On an ICAO procedure, you subtract 0.8 percent from the published gradient. A
7.6 percent gradient, for example, becomes 7.6 - 0.8 = 6.8 percent.
Figure: EFB-Pro Screen Grab, from Eddie's notes.
There are commercial products available that automate this process, such as CAVU
Companies EFB-Pro. This program is available for many aircraft types and includes a
database of airport runways. It does not, however, have a database of obstacle
departure procedures. The program does what it advertises it will do but, in my opinion,
there can be confusion behind what the numbers mean and the inherent risks.
When entering the obstacle departure procedure restrictions, you are asked to note if
the procedure uses the ICAO or TERPS standard. If you select ICAO, the program
subtracts 0.8 percent, the Minimum Obstacle Clearance, from what it calls the "Gross
Gradient" to produce a "Net Gradient." Likewise, if you select TERPS, the program
multiplies the value by (1 - 0.24), the TERPS Required Obstacle Clearance.
This can be confusing. TERPS and ICAO Doc 8168 do not use the terms "Gross
Gradient" and "Net Gradient" to describe ROC and MOC. Using this program your climb
gradient ends up being equal to the obstacle height plus the net takeoff flight
path factor. In our example above, you cross abeam the controlling obstacle by 219
feet, provided you don't encounter a tailwind or temperature inversion, and provided you
do everything by the book to that point.
That being said, using EFB-Pro you can load your aircraft to a higher weight since you
are reducing the climb gradient by the MOC or ROC. In our original G450 performance
problem, EFB-Pro allows us to load up to 55,720 lbs an increase of 7,493 lbs over what
the aircraft's FMS computed. (That makes sense, since the FMS only subtracts the net
takeoff flight path 0.8 percent factor, where as EFB-Pro subtracts the larger TERPS 24
percent ROC.)

Pros
There are many advantages to reducing your obstacle climb gradient by the TERPS
Required Obstacle Clearance (ROC) of 24 percent or the ICAO Minimum Obstacle
Clearance (MOC) of 0.8 percent:

1. Reducing the published gradient by the ROC or MOC still provides multi-engine
aircraft with their 14 CFR 25.115 "net takeoff flight path" margin, which is 0.8 percent
for two engine aircraft, 0.9 percent for three engine aircraft, and 1.0 percent for four
engine aircraft.
2. Using a software application can be more accurate than manually chasing a pencil
through hard-to-use charts.
3. Reducing the published gradient by the ROC or MOC will significantly increase the
available payload for departure, and this can mean the difference between going or
not going, or reaching one's destination or having to make a fuel stop.

Cons
The are also disadvantages:

1. Reducing your climb gradient by the ROC or MOC leaves you only with the 14 CFR
25.115 "net takeoff flight path" as a safety margin, and that may not be enough if the
 winds change from a headwind on the runway to a tailwind at altitude, or if there is a
temperature inversion. In our Aspen example, eliminating the 24 percent ROC
reduces your margin over the first major obstacle from 500 feet to just 219 feet, the
"net takeoff flight path" margin.
2. Chasing through the spaghetti charts can induce critical errors that are large enough
to eliminate the remaining safety margin. Using computer software can also be
subject to errors, since it is up to the user to input the correct gradient. Available
software does not include a database of the correct procedure gradients.
3. This method does nothing to ensure low, close-in obstacles are addressed.
4. It is up to you to ensure you can still meet the ODP climb gradient if you don't lose an
engine; your AFM may not be of much help here. (I don't think this is really a problem
but can't make that decision for you and your aircraft. For my technique for dealing
with this, see: Ensuring All-Engine-Operating (AEO) Performance at Higher Weights.)
5. The reduced gradient is still based on the wider lateral area of the obstacle departure
procedure construction (TERPS or ICAO Doc 8168). The increase in payload is often
insufficient to make a difference in a go / no-go decision.

Recommendation
I've used this method for many years after studying terrain charts and ensuring I fully
understood where the threat was and even then, only if I had reasonable confidence the
winds at altitude were not reversed. The best candidates were places in a large valley
where I knew I would be able to navigate away from the terrain no matter the weather.
There was an accident many years ago, where an Air Force turboprop was unable to
out-climb a mountain as result of a temperature inversion; so I added that to my list of
things to worry about. Nevertheless, I continued to use the method until something
better came along. And that leads us to . . .

A High Tech Strategy: A Terrain Data Base with Obstacle

Analysis
We have already seen that reducing the vertical Obstacle Departure Procedure (ODP)
climb gradient by the ICAO 0.8 percent Minimum Obstacle Clearance (MOC) or the
TERPS 24 percent Required Obstacle Clearance (ROC) can pay dividends in allowing
us to increase our payload while still providing adequate obstacle clearance in the event
of an engine failure. (See above: A Low Tech Strategy: Reduce the Vertical Margins.)
But this approach proves worrisome because it requires a number of steps on a chart or
with a piece of software that can subject to user errors and it doesn't address low, close-
in obstacles.
We have also seen that the lateral clearances afforded by all TERPS and ICAO
obstacle departure procedures is very generous and that if we are able to navigate to
tighter tolerances we can eliminate consideration of some obstacles that could be as
much as 3 nautical miles left or right of course centerline. If your airplane can navigate
to tighter tolerances, you should be able to eliminate the far off obstacles and reduce
your required climb gradient even further. You are permitted to do this in accordance
with U.S. AC 120-91 and ICAO Annex 6, Part I, ATT C. (See U.S. AC 120-91 (Lateral).)
Let's go back to our Aspen example. Recall that using the The Old School Strategy we
were limited to a takeoff grossweight of 48,227 lbs using the G450 FMS which
automatically reduces the entered ODP climb gradient by the net takeoff flight
path factor of 0.8 percent for a two-engine aircraft. Recall also that using the A Low
Tech Strategy: Reduce the Vertical Margins and the computer application EFB-Pro we
were able to reduce the climb gradient by the larger TERPS 24 percent ROC which
yielded a much higher takeoff grossweight of 55,720 lbs. Neither of these solutions
addressed the low, close-in obstacle problem and both retained the very wide lateral
obstacle consideration inherent in any TERPS obstacle departure procedure. What if we
could narrow the obstacle consideration zone to no more than the 3,000 feet lateral
zone required by U.S. AC 120-91? Recall finally that the on-course obstacles were only
at a gradient of 2.33 percent, substantially less than the 5.75 percent gradient just one
mile right of course. That reduction to 2.33 percent makes a big difference:

Figure: G450 Net gradient second segment climb, flaps 20, 2.33 percent, from Eddie's
notes.
For the purpose of illustration, if we assume there are no obstacles higher than those
found on course within 3,000 feet left or right of course, we can chase through the chart
to see our takeoff grossweight goes way up to over 70,000 lbs. I've done this only to
illustrate the potential. It would take a very large database of all obstacles and a very
capable computer to compare each of these to our route of flight to do the job properly.
Fortunately there are commercial vendors who have just this capability.

Aircraft Performance Group (APG) "Runway Analysis" Services

Aircraft Performance Group offers what they call "Runway Analysis" to evaluate
declared distances, runway slope, weather, and departure obstacles. Their services are
available on web-based or stand alone applications and through many flight planning
service providers. The departure obstacle analysis program plans for performance that
avoids every obstacle laterally or vertically, including low, close-in obstacles. In the
event of an engine failure the scheduled performance keeps you away from obstacles
by required ICAO and U.S. FAA minimums. This method is not perfect, but it is nearly
so.
Back to our Aspen example. Recall that using the A Low Tech Strategy: Reduce the
Vertical Margins and the computer application EFB-Pro we were able to reduce the
obstacle departure procedure's climb gradient by the TERPS 24 percent ROC which
yielded a much higher takeoff grossweight of 55,720 lbs. This gradient keeps us clear
laterally of every obstacle in the very wide TERPS obstacle clearance area. The APG
solution narrows our lateral obstacle distance to the 3,000 feet lateral zone required by
U.S. AC 120-91 and produces OEI gross weights that are much higher:
Figure: APG KASE data, runway 33, G450, flaps 20, 30 Jan 2016.
The APG data shows we can depart under the same conditions at 69,279 lbs, slightly
less than our attempt to manually derive on course obstacles. That makes sense
because the APG data casts a wider net than our on course exercise. What about the
"Requires use of attached special departure procedures" note attached to our selected
33DP column?

Figure: Example APG KASE special departure procedures 33DP, runway 33, G450,
flaps 20, 30 Jan 2016.
Reading the verbiage provided with the APG data, we see these instructions precisely
mimic the SARDD THREE procedure. (Shown at the top of this page.) In fact, it is more
precise, offering bank angles, a turn based on position and not altitude, and a specific
time to begin flap retraction and acceleration. We can, as a result, have confidence that
we can load our G450 to 69,279 lbs. and:

1. be able to stay clear of all obstacles in the event of an engine failure if we stay within
3,000 feet of our filed and planned course,
2. not have to worry about changing departure procedures in the event of an engine
failure,
3. be able to meet our obstacle departure procedure climb gradient if we don't have an
engine failure (because we've checked the "double the thrust" theory in the simulator,
see Ensuring All-Engine-Operating (AEO) Performance at Higher Weights),
4. have enough fuel to make it to our destination on the east coast, and
5. avoid all low, close-in obstacles.

Note, however, that there are more special procedures listed than just the DP1 that
follows the ground track of the SARDD THREE . . .

Figure: Example APG KASE special departure procedures 33DP5, runway 33, G450,
flaps 20, 30 Jan 2016.
It is unclear why you would select the special procedure 33DP5 other than, perhaps,
being able to fly from waypoint to waypoint is appealing. This flight track is not
duplicated by any of the airport's published procedures and it is doubtful you could file it
as a matter of normal operations. I suppose that once you've lost the engine and
declared an emergency you can fly any track you want and in some cases these APG
hybrid procedures will result in further increased departure grossweight. (It does not in
our Aspen example.) But if you plan using one of these unpublished procedures there is
something you need to factor into the equation . . .
 Figure: Comparing the KASE SARDD THREE to the APG KASE 33DP5, from Eddie's
notes (using Google earth).
The ground track of this procedure is different than the published SARDD THREE and
you will not be able to file it for your planned departure. That leaves you in the situation
where your FMS, ATC, and your departure briefing are all based on one thing (the
SARDD THREE) and your plan in the event of an engine failure is another thing entirely.
Are you going to have the presence of mind to make these changes after an engine
failure? Even if all that is involved is changing the active flight plan in the FMS and
making a radio call, you may have your hands full and this is just an unneeded
nuisance.

Pros
There are many advantages to using a computer application that melds a digital
obstacle file and terrain maps to narrow the obstacle clearance area with aircraft
performance data to produce a climb gradient reduced by the TERPS Required
Obstacle Clearance (ROC) of 24 percent or the ICAO Minimum Obstacle Clearance
(MOC) of 8 percent:
1. Reducing the published gradient by the ROC or MOC still provides multi-engine
aircraft with their "net takeoff flight path" margins, which are 0.8 percent for two
engine aircraft, 0.9 percent for three engine aircraft, and 1.0 percent for four engine
aircraft.
2. Narrowing the obstacle clearance area to no more than 3,000 feet laterally accounts
for all obstacles the aircraft is likely to encounter with modern navigation capability.
3. Accounting for all low, close-in obstacles allows the computed grossweight to out-
climb these obstacles if possible, or identify needed procedures to avoid them
laterally.
4. Reducing the published gradient by the ROC or MOC and narrowing the obstacle
clearance area will significantly increase the available payload for departure, and this
can mean the difference between going or not going, or reaching one's destination or
having to make a fuel stop.

Cons
The are two disadvantages that I can think of:

1. Planning takeoff grossweight based on an unpublished procedure can introduce

unwanted crew distractions when trying to quickly change FMS and other cockpit
settings following an engine failure as well as the need to let ATC know your ground
track is about to change. These procedures may have never been flight or simulator
tested, unlike the published procedures.
2. Pilots may not understand the mechanics involved when significantly increasing
takeoff grossweight and how that impacts the margin of safety and increases the
need to navigate more precisely. In our Aspen example we managed to increase our
departure grossweight by almost 25 percent but cut our vertical clearance over the
most restrictive obstacle by half. (An unexpected tailwind a few hundred feet in the
air could erase that margin entirely.) Our lateral tolerance after the first two turns
dropped from over 2 miles to less than a half of one mile.

Recommendation
I've been using APG data for over 10 years now and have had several issues, not the
least of which are those unpublished procedures. But I've become comfortable using it
and allow members of my flight department to use it provided they understand the
trade-offs and follow the precautions which I am about the outline right now . . .

Recommendations
You can safely use all three of these methods, provided you take the right precautions
and understand the trade-offs. My preference is as follows:

1. The "High Tech Strategy" (terrain data base and obstacle analysis software) — I
always default to running an APG analysis at any airport I am going to for the first
time.
 o If the resulting numbers default to the aircraft's climb limit, I know there isn't an
obstacle problem and I can then count on my aircraft's FMS performance
computer to keep me safe. If the numbers are reduced because of obstacles, I
use only published procedures and ensure I load my aircraft to a weight less
than or equal to the APG stated limit.
o In either case, I know I've solved the low, close-in obstacle problem as well.
o I also check the weather to ensure there is no chance of a tailwind during the
departure procedure or a significant temperature inversion.
o I will brief the crew that we are reducing our vertical and lateral safety margins
and the need to navigate within a half-mile is critical. Because of this, we need
to ensure our GPS RAIM is good, we need to set the RNP alert value on our
avionics to 0.25 to give us ample warning if we are about to venture out of the
3,000 feet lateral protection area, and we need to do a check of all this just
prior to takeoff.
o If I didn't have access to a program like APG's runway analysis, I would move
on to the next option . . .
2. The "Old School Strategy" (using One-Engine-Inoperative performance against the
published All-Engines-Operative obstacle departure procedure) — My next choice
would be to revert to using OEI performance against the AEO procedure with one
extra allowance. If flying a TERPS procedure, I will allow the G450 performance
computer's default 0.8 percent net takeoff flight path gradient reduction to determine
a lower gradient. This gives me a little extra vertical margin. If I have reduced my
vertical margin, I take the following precautions:
o I will try to understand the low, close-in obstacle problem by studying the
airport notes and if they are a threat, I will reduce my grossweight to a point
that guarantees I will cross the DER at least 200' above the DER. That
eliminates the problem. (Here's how to do that: How to out-climb all low, close-
in obstacles.)
o I also check the weather to ensure there is no chance of a tailwind during the
departure procedure or a significant temperature inversion.
o I will brief the crew that we are flying with a reduced vertical margin. Because
of this, we need to ensure our GPS RAIM is good, we need to set the RNP
alert value on our avionics to 0.25 to give us ample warning if we are about to
venture out of the 3,000 feet lateral protection area, and we need to do a
check of all this just prior to takeoff.
o If I was flying an airplane without the G450's ability to automate this process, I
would move on to the next option . . .
3. The "Low Tech Strategy" (reducing the AEO departure procedure by the full value of
0.8 percent MOC on an ICAO procedure or 24 percent ROC on a TERPS procedure)
— My last choice would be to use EFB-Pro because, in my opinion, the program is
not user friendly (and that increases the chance of an input error) and requires the
obstacle departure procedure be entered manually, introducing another opportunity
for error. That being said, I would only use it on a TERPS procedure and then I would
select the ICAO option to increase my vertical margin. From this point the
precautions are identical to Option Two.
I've made these choices because APG is provided at no cost with our flight planning
service provider (ARINCDirect) and the G450 has a very easy to use performance
computer built into the normal performance and takeoff data pages of the FMS. If you
have access to APG but your aircraft does not have a performance computer with this
capability, you may opt to elevate the "low tech" strategy as your Option Two. I can't
answer for you, but these are the choices I have made. No matter your choice, I
recommend you understand what margins you are cutting (vertical and/or lateral), the
need to navigate precisely and make the necessary GPS RAIM or other navigation
accuracy checks, and keep an eye on winds and temperature at altitude. I've used
these techniques for fifteen years now and they have often made the difference
between being able to takeoff or having to wait for the weather to improve.

Gray Area
We often think of gray areas as something with no right or wrong answer, but that isn't
right. I think a gray area is more than likely a problem you haven't thought through. I
often get asked "How can I be sure I'm going to clear an obstacle if I lose an engine
after I've retracted my flaps and have accelerated above V2 to V2+10 when the charts
are based on losing an engine at V1, climbing out with the flaps set at those speeds? It
gets worse. What if you do all that and lose an engine after your flaps are up?

Figure: Obstacle clearance, three scenarios, from Eddie's notes.

In the drawing there are four scenarios, two of which guarantee obstacle clearance and
two which leave you with question marks:
1. Blue line — If you load your airplane up to assure obstacle clearance One Engine
Inoperative (OEI), climb with your flaps set as planned, and fly at the recommended
speed (i.e., V2 to V2+10), you will be okay if you fly with the required precision.
2. Green solid line — If you load your airplane up to assure obstacle clearance One
Engine Inoperative (OEI) or All Engines Operating (AEO) using the appropriate
planning, climb with your flaps set as planned, and fly at the recommended speed
(i.e., V2 to V2+10), you will be okay if you fly with the required precision even if you
lose an engine at any point from V1 all the way through obstacle clearance.
3. Green dashed line — If you took the steps to assure obstacle clearance under
scenarios 1 or 2 above, but after you takeoff decide you are going to accelerate
above the target speed and clean up the airplane, you don't know if you are going to
clear the obstacle, even if you don't lose an engine.
4. Red line — If you took all the steps to assure obstacle clearance under scenarios 1
or 2 above, but after you takeoff decide you are going to accelerate above the target
speed and clean up the airplane, and then you lose an engine, you are even more
trouble. The best you can do is pull the nose up to catch the appropriate engine-out,
flaps up speed (i.e., VSE). Even if you do this, you don't know if you are going to clear
the obstacle.

This isn't so much a gray area as a pilot error in understanding what it takes to clear an
obstacle. The only way to assure obstacle clearance is to fly the target speed in the
correct configuration until the obstacles are cleared. Does that mean you need to
subject your passengers to a rocket ship ride even if you don't lose an engine on the
runway? Not necesarily. If you understand where the obstacles are you can adjust your
clean up altitude appropriately.
In our Aspen Example, above, we know the required climb gradient is dictated by two
obstacles at 4.5 and 7.2 nm north of the airport, the higher of which is 9,250' MSL
(1,570' above the departure end of the runway). We should plan on maintaining V 2 to
V2+10 until we've climbed above this altitude, even if we don't lose an engine. Now we
know we can beat the obstacle following an engine failure at V1, at our planned level off
altitude, or anywhere in between.

Complete Extracts of Cited Sources

It has taken me years to digest all of this and if you really want to understand the
process, there is no better way than to start reading. Be careful, however. The rules and
regulations from one source may draw you to a conclusion that is made unworkable by
the next. You need to read all of it.

14 CFR 25.111 Takeoff Path

[14 CFR 25, §25.111 Takeoff path.
(a) The takeoff path extends from a standing start to a point in the takeoff at which the
airplane is 1,500 feet above the takeoff surface, or at which the transition from the
takeoff to the en route configuration is completed and VFTO is reached, whichever point
is higher. In addition—
(2) The airplane must be accelerated on the ground to VEF, at which point the critical
engine must be made inoperative and remain inoperative for the rest of the takeoff.

14 CFR 25, §25.115

[14 CFR 25, §25.115] Takeoff flight path.
(a) The takeoff flight path shall be considered to begin 35 feet above the takeoff surface
at the end of the takeoff distance determined in accordance with §25.113(a) or (b), as
appropriate for the runway surface condition.
(b) The net takeoff flight path data must be determined so that they represent the actual
takeoff flight paths (determined in accordance with §25.111 and with paragraph (a) of
this section) reduced at each point by a gradient of climb equal to—
(1) 0.8 percent for two-engine airplanes;
(2) 0.9 percent for three-engine airplanes; and
(3) 1.0 percent for four-engine airplanes.
(c) The prescribed reduction in climb gradient may be applied as an equivalent
reduction in acceleration along that part of the takeoff flight path at which the airplane is
accelerated in level flight.

14 CFR 91, §91.13

[14 CFR 91, §91.13 (a)] Aircraft operations for the purpose of air navigation. No person
may operate an aircraft in a careless or reckless manner so as to endanger the life or
property of another.

14 CFR 91, §91.175 (f)

[14 CFR 91, §91.175 (f)] Civil airport takeoff minimums. This paragraph applies to
persons operating an aircraft under part 121, 125, 129, or 135 of this chapter.
[14 CFR 91, §91.175 (f)(3)] Except as provided in paragraph (f)(4) of this section, no
pilot may takeoff under IFR from a civil airport having published obstacle departure
procedures (ODPs) under part 97 of this chapter for the takeoff runway to be used,
unless the pilot uses such ODPs or an alternative procedure or route assigned by air
traffic control.
[14 CFR 91, §91.175 (f)(4)] Notwithstanding the requirements of paragraph (f)(3) of this
section, no pilot may takeoff from an airport under IFR unless:
[14 CFR 91, §91.175 (f)(4)(i)] For part 121 and part 135 operators, the pilot uses a
takeoff obstacle clearance or avoidance procedure that ensures compliance with the
applicable airplane performance operating limitations requirements under part 121,
subpart I or part 135, subpart I for takeoff at that airport; or
[14 CFR 91, §91.175 (f)(4)(ii)] For part 129 operators, the pilot uses a takeoff obstacle
clearance or avoidance procedure that ensures compliance with the airplane
performance operating limitations prescribed by the State of the operator for takeoff at
that airport.

14 CFR 135, §135.379

[14 CFR 135, §135.379] Large transport category airplanes: Turbine engine powered:
Takeoff limitations.
[14 CFR 135, §135.379 (d)] No person operating a turbine engine powered large
transport category airplane may take off that airplane at a weight greater than that listed
in the Airplane Flight Manual—
[14 CFR 135, §135.379 (d)(1)] For an airplane certificated after August 26, 1957, but
before October 1, 1958 (SR422), that allows a takeoff path that clears all obstacles
either by at least (35 + 0.01 D) feet vertically (D is the distance along the intended flight
path from the end of the runway in feet), or by at least 200 feet horizontally within the
airport boundaries and by at least 300 feet horizontally after passing the boundaries; or
[14 CFR 135, §135.379 (d)(2)] For an airplane certificated after September 30, 1958
(SR422A, 422B), that allows a net takeoff flight path that clears all obstacles either by a
height of at least 35 feet vertically, or by at least 200 feet horizontally within the airport
boundaries and by at least 300 feet horizontally after passing the boundaries.

AC 120-91, ¶ 10, 11, 12

[AC 120-91, ¶ 10.]
a. The Area Analysis Method defines an obstacle accountability area (OAA) within
which all obstacles must be cleared vertically. The OAA is centered on the intended
flight track and is acceptable for use without accounting for factors that may affect the
actual flight track relative to the intended track, such as wind and available course
guidance.
b. The Flight Track Analysis Method is an alternative means of defining an OAA based
on the navigational capabilities of the aircraft. This methodology requires the operator to
evaluate the effect of wind and available course guidance on the actual ground track.
While this method is more complicated, it can result in an area smaller than the OAA
produced by the Area Analysis Method.
 Figure: Straight Out Departures, from AC 120-91, Appendix 1, Figure 1.
[AC 120-91, ¶ 11.] Area Analysis Method
a. During straight-out departures or when the intended track or airplane heading is
within 15 degrees of the extended runway centerline heading, the following criteria
apply:
(1) The width of the OAA is 0.0625D feet on each side of the intended track (where D is
the distance along the intended flight path from the end of the runway in feet), except
when limited by the following minimum and maximum widths.
(2) The minimum width of the OAA is 200 feet on each side of the intended track within
the airport boundaries, and 300 feet on each side of the intended track outside the
airport boundaries.
(3) The maximum width of the OAA is 2,000 feet on each side of the intended track.
(See Appendix 1, Figure 1.)
On a straight out departure you have to be clear of all obstacles within an area that
starts out 200 feet on either side and ends up no wider than 2,000 feet on each side.
 Figure: Turning Departures, from AC 120-91, Appendix 1, Figure 2.
b. During departures involving turns of the intended track or when the airplane heading
is more than 15 degrees from the extended runway centerline heading, the following
criteria apply:
(1) The initial straight segment, if any, has the same width as a straight-out departure.
(2) The width of the OAA at the beginning of the turning segment is the greater of:
(a) 300 feet on each side of the intended track.
(b) The width of the OAA at the end of the initial straight segment, if there is one.
(c) The width of the end of the immediately preceding segment, if there is one, analyzed
by the Flight Track Analysis Method.
(3) Thereafter in straight or turning segments, the width of the OAA increases by 0.125D
feet on each side of the intended track (where D is the distance along the intended flight
path from the beginning of the first turning segment in feet), except when limited by the
following maximum width:
(4) The maximum width of the OAA is 3,000 feet on each side of the intended track.
(See Appendix 1, Figure 2.)
On a departure with a turn of more than 15 degrees, you have to be clear of all
obstacles within an area that starts out 200 feet on either side and ends up no wider
than 3,000 feet on each side.
c. The following apply to all departures analyzed with the Area Analysis Method:
(1) A single intended track may be used for analysis if it is representative of operational
procedures. For turning departures, this implies the bank angle is varied to keep a
constant turning radius with varying speeds.
(2) Multiple intended tracks may be accommodated in one area analysis by increasing
the OAA width accordingly. In a turn, the specified OAA half-widths (i.e., one-half of the
OAA maximum width) should be applied to the inside of the minimum turn radius and
the outside of the maximum turn radius. An average turn radius may be used to
calculate distances along the track.
(3) The distance to an obstacle within the OAA should be measured along the intended
track to a point abeam the obstacle.
(4) When an operator uses the Area Analysis Method, the operator does not need to
separately account for crosswind, instrument error, or flight technical error within the
OAA.
(5) Obstacles prior to the end of the runway need not be accounted for, unless a turn is
made prior to the end of the runway.
(6) One or more turns of less than 15 degrees each, with an algebraic sum of not more
than a 15 degree change in heading or track, may be analyzed as a straight-out
departure.
(7) No accountability is needed for the radius of the turn or gradient loss in the turn for a
turn with a 15 degree or less change in heading or track.

Flight Track Analysis Method

[AC 120-91, ¶ 12.] The Flight Track Analysis Method involves analyzing the ground
track of the flight path This paragraph discusses factors that the operator must consider
in performing a Flight Track Analysis.
a. Pilotage in Turns. The operator should consider the ability of a pilot to initiate and
maintain a desired speed and bank angle in a turn. Assumptions used here should be
consistent with pilot training and qualification programs.
b. Winds.
(1) When using the Flight Track Analysis Method while course guidance is not available,
operators should take into account winds that may cause the airplane to drift off the
intended track.
(2) The operator should take into account the effect of wind on the takeoff flight path, in
addition to making the headwind and tailwind component corrections to the takeoff
grossweight used in a straight-out departure.
(3) When assessing the effect of wind on a turn, the wind may be held constant in
velocity and direction throughout the analysis unless known local weather phenomena
indicate otherwise.
(4) If wind gradient information is available near the airport and flight path (e.g., wind
reports in mountainous areas adjacent to the flight path), the operator should take that
information into account in the development of a procedure.

ICAO Annex 6, Part I, ATT C

[ICAO Annex 6, Part I, ATT C, ¶1] The purpose of this Attachment is to provide
guidance as to the level of performance intended by the provisions of Chapter 5 as
applicable to turbine-powered subsonic transport type aeroplanes over 5,700 kg
maximum certificated take-off mass having two or more engines.
[ICAO Annex 6, Part I, ATT C-2] The terms “accelerate-stop distance”, “take-off
distance”, “V1”, “take-off run”, “net take-off flight path”, “one engine inoperative en-route
net flight path”, and “two engines inoperative en-route net flight path”, as relating to the
aeroplane, have their meanings defined in the airworthiness requirements under which
the aeroplane was certificated. If any of these definitions are found inadequate, then a
definition specified by the State of the Operator should be used.]
[ICAO Annex 6, Part I, ATT C, ¶2.8.1] The net take-off flight path is the one-engine-
inoperative flight path which starts at a height of 10.7 m (35 ft) at the end of the take-off
distance required and extends to a height of at least 450 m (1,500 ft) calculated in
accordance with the conditions of 2.9, the expected gradient of climb being diminished
at each point by a gradient equal to:

• 0.5 per cent, for aeroplanes with two engines,

• 0.8 per cent, for aeroplanes with four engines.

[ICAO Annex 6, Part I, ATT C, ¶5]

5.1 No aeroplane should commence a take-off at a mass in excess of that shown in the
flight manual to correspond with a net take-off flight path which clears all obstacles
either by at least a height of 10.7 m (35 ft) vertically or at least 90 m (300 ft) plus 0.125D
laterally, where D is the horizontal distance the aeroplane has travelled from the end of
take-off distance available, except as provided in 5.1.1 to 5.1.3 inclusive. For
aeroplanes with a wingspan of less than 60m (200 ft) a horizontal obstacle clearance of
half the aeroplane wingspan plus 60 m (200 ft), plus 0.125D may be used. In
determining the allowable deviation of the net take-off flight path in order to avoid
obstacles by at least the distances specified, it is assumed that the aeroplane is not
banked before the clearance of the net take-off flight path above obstacles is at least
one half of the wingspan but not less than 15.2 m (50 ft) height and that the bank
thereafter does not exceed 15°, except as provided in 5.1.4. The net take-off flight path
considered is for the altitude of the aerodrome and for the ambient temperature and not
more than 50 per cent of the reported headwind component or not less than 150 per
cent of the reported tailwind component existing at the time of take-off. The take-off
obstacle accountability area defined above is considered to include the effect of
crosswinds.
5.1.1 Where the intended track does not include any change of heading greater than
15°,
a) for operations conducted in VMC by day, or
b) for operations conducted with navigation aids such that the pilot can maintain the
aeroplane on the intended track with the same precision as for operations specified in
5.1.1 a), obstacles at a distance greater than 300 m (1,000 ft) on either side of the
intended track need not be cleared.
5.1.2 Where the intended track does not include any change of heading greater than
15° for operations conducted in IMC, or in VMC by night, except as provided in 5.1.1 b);
and where the intended track includes changes of heading greater than 15° for
operations conducted in VMC by day, obstacles at a distance greater than 600 m (2,000
ft) on either side of the intended track need not be cleared.
5.1.3 Where the intended track includes changes of heading greater than 15° for
operations conducted in IMC, or in VMC by night, obstacles at a distance greater than
900 m (3,000 ft) on either side of the intended track need not be cleared.
5.1.4 An aeroplane may be operated with bank angles of more than 15° below 120 m
(400 ft) above the elevation of the end of the take-off run available, provided special
procedures are used that allow the pilot to fly the desired bank angles safely under all
circumstances. Bank angles should be limited to not more than 20° between 30 m (100
ft) and 120 m (400 ft), and not more than 25° above 120 m (400 ft). Methods approved
by the State of the Operator should be used to account for the effect of bank angle on
operating speeds and flight path including the distance increments resulting from
increased operating speeds. The net take-off flight path in which the aeroplane is
banked by more than 15° should clear all obstacles by a vertical distance of at least
10.7 m (35 ft) relative to the lowest part of the banked aeroplane within the horizontal
distance specified in 5.1. The use of bank angles greater than those mentioned above
should be subject to the approval from the State of the Operator.

ICAO Annex 8, Part IIIA, ¶2.2.3

[ICAO Annex 8, Part IIIA, ¶2.2.3] Performance data shall be determined and scheduled
in the flight manual so that their application by means of the operating rules to which the
aeroplane is to be operated in accordance with 5.2 of Annex 6, Part I, will provide a safe
relationship between the performance of the aeroplane and the aerodromes and routes
on which it is capable of being operated. Performance data shall be determined and
scheduled for the following stages for the ranges of mass, altitude or pressure- altitude,
wind velocity, gradient of the take-off and landing surface for landplanes; water surface
conditions, density of water and strength of current for seaplanes; and for any other
operational variables for which the aeroplane is to be certificated. 2.2.3.1 Take-off. The
take-off performance data shall include the accelerate-stop distance and the take-off
path.
2.2.3.1.1 Accelerate-stop distance. The accelerate-stop distance shall be the distance
required to accelerate and stop, or, for a seaplane to accelerate and come to a
satisfactorily low speed, assuming the critical engine to fail suddenly at a point not
nearer to the start of the take-off than that assumed when determining the take-off path
(see 2.2.3.1.2).
2.2.3.1.2 Take-off path. The take-off path shall comprise the ground or water run, initial
climb and climb-out, assuming the critical engine to fail suddenly during the take-
off (see 2.2.3.1.1). The take-off path shall be scheduled up to a height that the
aeroplane can maintain and at which it can carry out a circuit of the aerodrome. The
climb-out shall be made at a speed not less than the take-off safety speed as
determined in accordance with 2.3.1.3.
2.3.1.3 Take-off safety speed. The take-off safety speeds assumed when the
performance of the aeroplane (after leaving the ground or water) during the take-off is
determined shall provide an adequate margin above the stall and above the minimum
speed at which the aeroplane remains controllable after sudden failure of the critical
engine.

ICAO Doc 8168 Vol II, ¶2

[ICAO Doc 8168 Vol II, ¶2.2.6] The standard procedure design gradient (PDG) is 3.3
per cent (Cat H, 5.0 per cent). The PDG begins at a point 5 m (16 ft) above the
departure end of the runway (DER).
Cat H, by the way, applies to helicopters.
[ICAO Doc 8168 Vol II, ¶2.2.7] The standard PDG provides an additional clearance of
0.8 per cent of the distance flown from the DER, above an obstacle identification
surface (OIS). The OIS has a gradient of 2.5 per cent (Cat H, 4.2 per cent).
[ICAO Doc 8168 Vol II, ¶2.2.8] Where an obstacle penetrates the OIS, a steeper PDG
may be promulgated to provide obstacle clearance of 0.8 per cent of the distance flown
from the DER.
[ICAO Doc 8168 Vol II, ¶2.5.1] The minimum obstacle clearance (MOC) in the primary
area is 0.8 per cent of the distance flown from the DER. The MOC is zero at the DER.
[ICAO Doc 8168 Vol II, ¶2.5.2] The MOC is provided above an obstacle identification
surface or, where an obstacle penetrates the OIS, above the elevation of the obstacle.
[ICAO Doc 8168 Vol II, ¶2.7.5] An increased gradient that is required to a height of 60 m
(200 ft) or less, (normally due to low, close-in obstacles) shall not be promulgated (see
Figure I-3-2-3). The position and elevation/height of close-in obstacles penetrating the
OIS shall be promulgated (see Chapter 5, “Published information for departure
procedures”).

ICAO Doc 8168 Vol II, §3, ¶1.7

1.7.1 The design of procedures in accordance with this section assumes normal
operations and that all engines are operating.
1.7.2 It is the responsibility of the operator to conduct an examination of all relevant
obstacles and to ensure that the performance requirements of Annex 6 are met by the
provision of contingency procedures for abnormal and emergency operations. Where
terrain and/or obstacle considerations permit, the contingency procedure routing should
follow that of the departure procedure.
1.7.3 It is the responsibility of the State to make available the obstacle information
described in Annexes 4 and 6, and any additional information used in the design of
departures in accordance with this Section.

ICAO Doc 8168 Vol II, ¶3.2.4.1

[ICAO Doc 8168 Vol II, ¶3.2.4.1] Departure with no track adjustment. The area begins at
the DER and has an initial width of 300 m (Cat H, 90 m). It is centred on the runway
centre line and splays at an angle of 15° on each side of the extended runway center
line (see Figure I-3-3-1). The area terminates at the end of the departure procedure as
specified in Chapter 2, 2.4, “End of the departure procedure.”
[ICAO Doc 8168 Vol II, ¶2.4] The departure procedure ends at the point where the PDG
reaches the minimum altitude/height authorized for the next phase of flight (i.e. en-
route, holding or approach).

TERPS Volume 1, ¶201

[TERPS, Volume 1]
201. TERPS. Concept of Primary Required Obstacle Clearance (ROC). The title of this
order, United States Standard for Terminal Instrument Procedures (TERPS), contains a
key word in defining the order's content. The word is "STANDARD;" something set up
and established by authority as a rule for the measure of quantity, weight, extent, value,
or quality.
a. The TERPS document specifies the minimum measure of obstacle clearance that is
considered by the FAA (the Federal authority) to supply a satisfactory level of vertical
protection. The validity of the protection is dependent, in part, on assumed aircraft
performance. In the case of TERPS, it is assumed that aircraft will perform within
certification requirements.
b. The following is an excerpt from the foreword of this order: "These criteria are
predicated on normal aircraft operations for considering obstacle clearance
requirements." Normal aircraft operation means all aircraft systems are functioning
normally, all required navigational aids (NAVAID's) are performing within flight
inspection parameters, and the pilot is conducting instrument instrument operations
utilizing procedures based on the TERPS standard to provide ROC. While the
application of TERPS criteria indirectly addresses issues of flyability and efficient use of
NAVAID's, the major safety contribution is the provision of obstacle clearance
standards. This facet of TERPS allows aeronautical navigation in instrument
meteorological conditions (IMC) without fear of collision with unseen obstacles. ROC is
provided through application of level and sloping OCS.
TERPS, Volume 1, ¶203
[TERPS, Volume 1, ¶203.] Sloping Obstacle Clearance Surfaces (OCS). The method of
applying ROC, in segments dedicated to descending on a glidepath or climbing in a
departure or missed approach segment, requires a different obstacle clearance concept
than the level OCS because the ROC value must vary throughout the segment. The
value of ROC near the runway is relatively small, and the value at the opposite end of
the segment is sufficient to satisfy one of the level surface standards above. It follows
then, that a sloping OCS is a more appropriate method of ROC application.
[TERPS, Volume 1, ¶203.b.] The concept of providing obstacle clearance in the climb
segment, in instrument procedures, is based on the aircraft maintaining a minimum
climb gradient. The climb gradient must be sufficient to increase obstacle clearance
along the flightpath so that the minimum ROC for the subsequent segment is achieved
prior to leaving the climb segment (see figure 1-3). For TERPS purposes, the MINIMUM
climb gradient that will provide adequate ROC in the climb segment is 200 ft/NM.
While this "MINIMUM climb gradient" of 200 feet per nautical mile is the same as the
ICAO 3.3 percent Procedure Design Gradient (PDG) and the 152 feet per nautical mile
Obstacle Clearance Surface is the same as the ICAO 2.5 percent Obstacle
Identification Surface, the similarities end there.
[TERPS, Volume 1, ¶203.b.(1)] The obstacle evaluation method for a climb segment is
the application of a rising OCS below the minimum climbing flightpath. Whether the
climb is for departure or missed approach is immaterial. The vertical distance between
the climbing flightpath and the OCS is ROC. ROC for a climbing segment is defined as
ROC = 0.24 CG . This concept is often called the 24 percent rule. Altitude gained is
dependent on climb gradient (CG) expressed in feet per NM. The minimum ROC
supplied by the 200 ft/NM CG is 48 ft/NM (0.24 x 200 = 48). Since 48 of the 200 feet
gained in 1 NM is ROC, the OCS height at that point must be 152 feet (200 - 48 = 152),
or 76 percent of the CG (152 ÷ 200 = 0.76). The slope of a surface that rises 152 over 1
NM is 40 (6076.11548 ÷ 152 = 39.97 = 40).
[TERPS, Volume 1, ¶203.b.(2)] Where an obstruction penetrates the OCS, a
nonstandard climb gradient (greater than 200 ft/NM) is required to provide adequate
ROC. Since the climb gradient will be greater than 200 ft/NM, ROC will be greater than
48 ft/NM (0.24 x CG > 200 = ROC > 48). The nonstandard ROC expressed in ft/NM can
be calculated using the formula: (0.24 h) ÷ (0.76d) where "h" is the height of the
obstacle above the altitude from which the climb is initiated, and "d" is the distance in
NM from the initiation of climb to the obstacle. Normally, instead of calculating the
nonstandard ROC value, the required climb gradient is calculated directly using the
formula: h ÷ (0.76d).
[TERPS, Volume 1, ¶203.c.] In the case of an instrument departure, the OCS is applied
during the climb until at least the minimum en route value of ROC is attained. The OCS
begins at the departure end of runway, at the elevation of the runway end. It is assumed
aircraft will cross the departure end-of-runway at a height of at least 35 ft. However, for
TERPS purposes, aircraft are assumed to lift off at the runway end (unless the
procedures state otherwise). The ROC value is zero at the runway end, and increases
along the departure route until the appropriate ROC value is attained to allow en route
flight to commence.

TERPS, Vol 4, ¶ 1.3.1

Do not publish a CG to a height of 200 feet or less above the DER elevation. Annotate
the location and height of any obstacles that cause such climb gradients.

TERPS, Vol 4, ¶1.4.6

Where obstacles 3 statute miles or less
􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉􏱉from the DER
penetrate the OCS:
(1) Publish a note identifying the obstacle(s) type, location relative to DER, AGL height,
and MSL elevation, and
(2) Publish standard takeoff minimums with a required CG to a specified altitude, and
(3) Publish a ceiling and visibility to see and avoid the obstacle(s), and/or
(4) Develop a specific textual or graphic route to avoid the obstacle(s).
NOTE: Where low, close-in obstacles result in a climb gradient to an altitude 200 feet or
less above DER elevation, only paragraph 1.4.6a(1) applies.

TERPS, Volume 4, ¶1.6

• The ICA is an area centered on the runway centerline extended used to evaluate
obstacle clearance during the climb to 400 feet above DER (minimum climb
gradient 200 ft/NM).
• The Initial Climb Area Baseline (ICAB) is a line extending perpendicular to the
runway centerline 􏱉 500 at DER. It is the origin of the ICA (see figure 1-5).
• The Initial Climb Area End-line (ICAE) is a line at the end of the ICA
perpendicular to the runway centerline extended. The splay of 15° and length of
the ICA determine its width (see figure 1-5).
• The ICA length is normally 2 NM, measured from the ICAB to the ICAE along
runway centerline extended. It may be less than 2 NM 􏱉in length for early turns
􏱉by publishing a climb gradient, or a combination of climb gradient and reduction
in TORA. The ICA may be extended beyond 2 NM 􏱉to maximum length of 10
NM. A specified altitude (typically 400' above DER) or the interception of [Positive
Course Guidance] (PCG) route must identify the ICAE.
• The ICA origin is 1,000 feet (􏱉500 perpendicular to runway centerline) wide at
the DER. The area splays outward at a rate of 15° relative to the departure
course (normally runway centerline).

References
14 CFR 25, Title 14: Aeronautics and Space, Airworthiness Standards: Transport
Category Airplanes, Federal Aviation Administration, Department of Transportation
14 CFR 91, Title 14: Aeronautics and Space, General Operating and Flight Rules,
Federal Aviation Administration, Department of Transportation
14 CFR 97, Title 14: Aeronautics and Space, Standard Instrument Procedures, Federal
Aviation Administration, Department of Transportation
14 CFR 121, Title 14: Aeronautics and Space, Operating Requirements: Domestic,
Flag, and Supplemental Operations, Federal Aviation Administration, Department of
Transportation
14 CFR 125, Title 14: Aeronautics and Space, Certification and Operations: Airplanes
Having a Seating Capacity of 20 or More Passengers or a Maximum Payload Capacity
of 6,000 Pounds or More; and Rules Governing Persons on Board Such Aircraft,
Federal Aviation Administration, Department of Transportation
14 CFR 135, Title 14: Aeronautics and Space, Operating Requirements: Commuter and
On Demand Operations and Rules Governing Persons on Board Such Aircraft, Federal
Aviation Administration, Department of Transportation
14 CFR 139, Title 14: Aeronautics and Space, Certification of Airports
Advisory Circular 120-91, Airport Obstacle Analysis, 5/5/06, U.S. Department of
Transportation
Aeronautical Information Manual
Gulfstream G450 Airplane Flight Manual, Revision 36, December 5, 2013
ICAO Annex 6 - Operation of Aircraft - Part 1 Commercial Aircraft, International
Standards and Recommended Practices, Annex 6 to the Convention on International
Civil Aviation, Part I, July 2010
ICAO Annex 8 - Airworthiness of Aircraft, International Standards and Recommended
Practices, Annex 8 to the Convention on International Civil Aviation, July 2010
ICAO Doc 8168 - Aircraft Operations - Vol II - Construction of Visual and Instrument
Flight Procedures, Procedures for Air Navigation Services, International Civil Aviation
Organization, 2006
United States Standard for Terminal Instrument Procedures (TERPS), Federal Aviation
Administration 8260.3B CHG 25, 03/09/2012