What’s the most overlooked and misunderstood part of IFR flying? I nominate the obstacle departure procedure (ODP). While almost any instrument pilot can recite trivia like holding pattern entries or VOR test requirements—important but relatively rare procedures—many are quite shaky on ODPs. That’s a shame, because ignorance of this procedure can be fatal.
A recent Bonanza crash is a sobering example. The instrument-rated pilot landed in the mountains of North Carolina as the sun was setting. After taking on fuel he filed an IFR flight plan and departed, but flew into a mountain just minutes after takeoff, killing himself and his passenger. As always, there are lots of important details here (the pilot had only owned the airplane for four months and the gear may have been left down on climb, raising questions about proficiency), but the key facts are not in doubt. This was a classic Controlled Flight Into Terrain accident.
From the preliminary NTSB report:
“Takeoff minimums and obstacle departure procedures for RHP (an uncontrolled airport) required pilots to remain within 3 nautical miles of the airport while climbing in visual conditions to cross the airport westbound at or above 4,900 ft mean sea level (msl). Then climb to 7,000 ft on a heading of 251° to the Harris (HRS) VORTAC 356° radial to HRS before proceeding on course. The procedure is not authorized at night.
Review of preliminary FAA Automatic Dependent Surveillance – Broadcast (ADS-B) data indicated that the pilot departed runway 8 and made a slight left turn toward the northeast. The last recorded data point showed the airplane about 3,750 ft, in a 656 ft-per-minute climb at 98 knots, on a course of 042°. The last ADS-B data point was located about 500 ft laterally from the initial impact with pine trees at an approximate elevation of 3,950 ft.
The RHP weather at 1945 included scattered clouds at 1,400 ft, broken clouds at 3,200 ft, and 7 miles visibility in rain. Sunset at Andrews was about 1917 and the end of civil twilight was about 1941.”
Clearly the Bonanza pilot did not follow the ODP, not even a little bit, yet the weather conditions definitely required some type of initial route other than direct—it is quite unlikely he could have climbed in visual conditions to 4900 feet as noted. Did he not know about the ODP or did he simply ignore it? We’ll never know, but both are plausible answers.
Flying an airplane under instrument flight rules is a privilege, and with that comes the responsibility to understand each part of the flight, with its accompanying rules and requirements—something the Bonanza pilot did not do. But the list of rules and requirements can be long and intimidating, from alternate airport criteria to fuel requirements to reporting points, and instrument departures don’t get the attention they deserve. As the FAA’s Instrument Procedures Handbook states, the burden is on the pilot in command:
“It is the pilot’s responsibility to determine if there is an ODP published for that airport. If a Part 91 pilot is not given a clearance containing an ODP, SID, or radar vectors and an ODP exists, compliance with such a procedure is the pilot’s choice.”
This leaves ODPs in a gray area. Following the procedure is definitely a good idea at night or in marginal weather conditions, but you don’t have to fly it and you won’t be cleared for it. There are very few areas where the FAA gives pilots so much latitude. Compared to the rigid world of instrument approaches—fly to this waypoint at this altitude, then descend only along this course—departures are almost shockingly relaxed. Essentially Air Traffic Control is saying, “do whatever you want until you climb into radar contact… and try not to hit anything.”
ODPs are so overlooked in part because they are for the convenience of the pilot, not ATC. That is the exact opposite of Standard Instrument Departures (SIDs), which exist to help ATC move high performance airplanes into the en route system. SIDs serve busy airports and are used by airlines on almost every flight, so they are charted and part of a clearance; ODPs serve smaller airports, so they are typically not charted and rarely heard on the radio. That doesn’t make them any less important.
So while the blame ultimately lies with the pilot in the accident above, there is no reason why ODPs should be so mysterious. If we are serious about making flying safer and more available to the public, we have to do more than just tut-tut about “pilots these days…” or say “he should have known better.”
An important first step would be to make ODPs more visible, so pilots are forced to consider them as a part of preflight planning. This could take many forms, like moving the textual descriptions from the front of FAA approach plate books, where they are easily skipped, to a page next to each airport’s other instrument procedures. A more logical flow to the charts—airport diagram, departure procedures, arrivals, approaches—would encourage good habits. Approach charts and airport diagrams could also have a bigger warning about any ODPs that exist, instead of the tiny symbol. Jeppesen charts are a little better organized in this respect; FAA charts need to close the gap.
In the app world, the options are even more attractive. ForeFlight recently added a small mountain symbol to alert pilots when high terrain is within 10 miles of an airport, a helpful addition but one that’s easy to miss. When planning a route on the Maps or Flights page, where most pilots spend their time, there is no suggestion that an ODP exists. The one-touch Procedure or Route options are tremendously helpful tools, but if you plan a flight from a mountain airport you won’t see any recommended routes that include a circling climb. A proactive suggestions would be a major improvement.
Perhaps the best change would be to add graphical depictions of ODPs whenever possible, like the charts available for traditional SIDs. Charted ODPs exist, and are growing as new RNAV ODPs are created, but they are still relatively rare. This limitation may have made sense in the days of paper charts and limited pages per book, but in 2021 such concerns are much less important. Even if the FAA does not publish formal charts for ODPs, apps could show suggested flight paths. For example, ForeFlight can draw a traffic pattern entry; it would be just as helpful to draw ODPs.
Controllers could also help, by adding ODPs to the clearances they read. Even a simple reminder that such a procedure exists might have prevented that Bonanza pilot from turning on course too soon (“check ODP”).
You might think it foolish to hope for the FAA to change ATC procedures, but there is plenty of precedent. After famed test pilot Scott Crossfield flew his Cessna 210 into a thunderstorm and crashed, ATC updated its procedure and now tells pilots about all convective weather ahead. Most of the time the pilot knows very well about thunderstorms along their route of flight, but it doesn’t hurt to confirm. “Trust but verify” is the unofficial motto of the pilot-controller relationship.
An even more significant change affected ground operations. Over the last two decades the FAA has pushed hard to reduce runway incursions, and one key initiative involved how taxi clearances are issued by ground controllers.
Consider how much clearer the instructions are now:
- Old way: “Taxi runway 3R.” Are you cleared to cross any runways you encounter along the way? Which taxiways should I use? As a student pilot, I was very unsure.
- New way: “Taxi to runway 3R via Alpha and Charlie, cross runway 7, hold short runway 3L.” This requires a few more words, but the instructions are crystal clear.
Obviously not every airport has an obstacle departure procedure—only if a close-in obstacle is present would one be created, and then only if the airport has an instrument approach—but the airports with ODPs are precisely the ones where pilots can get into trouble. It’s time to promote ODPs into the big leagues, along with SIDs, STARs, and approaches.
- The truth about learning to fly - May 15, 2023
- What it means to fly like a pro: 12 habits - May 12, 2023
- Go or No Go: spring cold front - April 12, 2023
Wholeheartedly agree. Well said.
Agree absolutely with Cameron. It is stunning that ODPs not given much more prominent exposure in FAA workproducts.
The fact that Part 91 operators are not required to comply with Take Off Minimums contributes to the lack of awareness of ODPs. The US happens to be one of the few countries that allows this. Pilots tend to disregard Take Off Minimums because they are not required to comply with them. Take Off Minimum requirements are a good heads up that an ODP is required. The Take Off Minimums for KRHP are 3,400’ and 2 miles.
This discussion of ODPs is long overdue. Especially considering they outnumber SIDs by an order of magnitude! That is probably why they appear in text format only. It isn’t practical to represent them graphically and expect (superannuated) pilots to fly with a phone book on their lap. Full conversion to electronic plates should fix that.
Most ODPs are designed conservatively allowing pilots and controllers to ignore them when convenient. That expedites flights but also engenders a lack of respect for the procedures as a class. I don’t know how to fix this. We may already be near an optimum system.
John Bone suggested checking the approach chart for take off minimums to see whether the field is likely to have an ODP. To that I’d add checking the circling minimums and possibly the MSA.
A quick look at the vfr sectional map shows how important the southwest climb out is for that airport.
A quick look at the chart should be first and foremost for any airport, and any direction, regarding approaching or departing.
That’s my home airport. Great day, VFR airport. Horrible instrument airport. Down in the end of a valley. LP is best approach with MDA almost 2300ft AGL. That ought to tell you something before you ever look at ODP on how to get out. Almost 20degree offset of runway to final approach course and centerline of runway runs you right into the mountains. NA at night. You have to get down 2300ft in 1/2 mile if you are at MDA and MAP. Missed approach turns you right into mountains where Bonanza hit. This is an airport where local knowledge is everything for instrument work. I’ve never seen anybody local use the ODP. Ever. In 3 mile radius you already hit mountains. Nobody is going to take off runway 8 if going instrument. Nobody likes to do runway 8 even if VFR. Everyone will take off runway 26, even if somewhat downwind. When you break ground you turn to 241 (not 251 as ODP says) and it takes you down the valley into wide open space. I’ve tried both the ODP and missed approach on clear days and they will both scare you to death. Weather on that day was horrible. Bonanza should have gone to KTYS in Knoxville, KAVL Asheville or one of the myriad small airports away from the mountains. I agree that a look at ODPs are long overdue. That should start with making sure the ODPs make sense.
Great points Chip. Local knowledge is everything…..mountain flying is unforgiving (including the “small” east coast ranges)….should be a daytime VFR sport only, especially for piston drivers. If I am ever considering an unknown airport in mountainous areas, I always reference the instrument approach procedures as they will give you local knowledge without being a local.
I drove past Andrews airport many times on highway 19 as a young man. You have to be there to fully appreciate the hazard of terrain.
I agree the airport is an extremely poor candidate for an ODP. I’ve rarely seen one that needed a speed restriction to minimize the turning radius in the initial climb. I’ve NEVER seen one that specified speed restrictions for a CIRCLING climb. It makes me wonder whether the FAA created this ODP in response to pressure from a previous CFIT mishap.
Given the conditions, I’m puzzled the pilot did not have the navigator, or one of their other devices set on the terrain page.
Missed approach turns you toward the mountains, not into the mountains as I see it. The terrain lines show the area at 3000′ while you should already be at about 4000″ on the missed and climbing.
Hi Erik. LP MDA is 3980ft and runway is 1699ft. No way you can get down 2300ft in 1/2 mile from MAP. You will have needed to make contact with runway environment from somewhere out closer to FAF to make stabilized approach. Then if you hit a layer of scud and lose contact with the runway and need to shoot the missed you are in real trouble. I have practiced the missed approach in VFR just to see what it’s like and there’s not that much clearance. Even referencing the instrument plate hardly tells the whole story (at least plainly). If you look at it you would think the airport is in the middle of the valley. It’s not. You might think that because the missed approach turns to the left that a takeoff on runway 8 and turn to the left would be a great way to start the ODP. It’s not. If you look on the instrument procedure right off the end of runway 8 is an obstruction at 2420ft and a bit off the runway 26 end an obstruction 2435. These are not tall towers. They’re not even shown on the sectional. They are maybe 50-100ft. Point is that you are already into rising terrain immediately adjacent to the airport. Traffic pattern on runway 8 is right but even VFR you had better be keeping it tight. We typically fly our pattern a couple of hundred feet higher than normal just because. When you are on base for 8 or 26 you are looking into peoples front porch. Local knowledge here. Back to the original point of ODPs, this one is horrible. I think they would have been better off to fly the approach backwards. However, you typically can’t make contact with ATC until you are around 5000ft. I tried filing with the ODP once and my clearance was direct to HRS. Perhaps they shouldn’t even have a ODP here.
Excellent summary and recommendations, John. As a CFI/CFI-I, I wholeheartedly agree that the ODPs should be easier to find, made graphical, and not buried in the Chart Supplement. Let’s prompt FAA to make those changes.
Victor, As I’m sure you know. the ODP’s are in the same booklet as the Instrument Approach Procedure Charts, not the Chart Supplements.
Several years ago, I requested Foreflight to add the option of choosing an ODP to be added to the route. My request was dismissed.
Completely agree with giving ODP same status as SIDs for charting and to include terrain avoidance plan query from ATC during clearance filing/delivery/initial controlling contact. This would also be a red flag for VFR departures that they should look twice.
Other big concern is the confusion on terrain clearance services from ATC. Just as ATC is very clear when announcing “no traffic separation services”, it should have a similar warning about “no terrain clearance services”. Absence of those services (and airspace warnings) is implied in flight following or when below “vectoring altitude” but I’d bet there’s a fair number that assume in ATC contact this is provided. More arcane is understanding the point when traffic, terrain, and airspace services are initiated during the sequence of ATC voice contact airborne, “radar contact” and updated clearance (or depending on airspace, if ever provided). San Diego bizjet departure into mountain comes to mind.
Well written and a timely topic!
I do undoubtedly agree with You, Mr. Jonh Zimmerman
Agree completely. In the last few years I transitioned from military and airline due to age-force retirement. I joined a great flying club and started flying light aircraft seriously for the first time ever. I planned a flight to KGKT and discovered an ODP at the bottom of the 10-9 page. After flying SIDs for decades to find something so important hidden in the fine print was shocking to me! Yes, let’s publish ODPs as departures, and Foreflight and the other apps can lead the way by placing ODPs more prominently. Great job Mr. Zimmerman.
I’m based at KGKT and file IFR almost every flight. While I always plan to depart on the ODP, recently I’ve been receiving it, “execute the ODP”, as part of my clearance from Knoxville Approach. This is an example of what John Z is suggesting, where ATC is taking an active role in ensuring pilots are aware of and execute ODPs. The relatively recent change in communication practice of listing the phone numbers for clearance delivery for uncontrolled fields makes this even easier, as a phone conversation tends to be better for most folks in inviting the opportunity to ask questions and receive clarification.
Awareness is the key.
I also agree with making the ODPs more prominent in both our printed and EFB publications so we have a better chance of knowing they are there.
Well written and whole heartedly agreed. ODP’s seem well-loved by those who know them, but even after so many years, it appears to be a poorly-taught subject. I’m honestly not sure why that is, because ever since my IFR training in the early 2000’s, I kept reading about how it was such a poorly-taught subject. You’d think it would’ve received some emphasis over the years.
Separately, you might want to edit the modern taxi phraseology, it generally doesn’t start with “taxi to,” any more. See FAA Order 7110.65Z 3-7-2 (b).
I wonder why in the age of GPS/RNP/PBN approaches SID’s, ODP’s etc. do not provide vertical guidance as well. You could climb out safely as you came in on the glide and would not need to worry about terrain.
The climb out procedure would have to state the minimum climb rate required while the pilot would need to consider his aircraft’s performance in view of possible downdrafts and tailwinds.
These kinds of departures could be flown on autopilot thus removing a lot of stress from the pilot and giving him the opportunity to keep full oversight of the aircraft in context with the airborne situation. In single pilot operations this can clearly get overwhelming when stress builds due to the need to handle multiple simultaneous operations that not always can be foreseen or planned.
Great article. It is so important to make changes in this area. I wholeheartedly support the recommendations.
I might add that the reported weather was below takeoff minimums. I realize that Part 91 flights do not have to abide by takeoff minimums , but they are put there for good reason.
John: An excellent insightful and timely analysis and proposal that needs to be presented to the FAA Administrator, the Chairs of the NTSB, House and Senate Transportation Committees and the CEOs of Foreflight and other similar apps —the folks that are in a position to provide leadership and have the power to make it happen. A supporting coalition of the alphabet groups would also be helpful and persuasive.
If this idea circulates on the desks of lower level bureaucrats, it is likely to languish there for years. A directive from the head shed to get it done is more likely to succeed!
Let’s go ForeFlight! You are the ground breakers. Now is the time to come to the aid of your pilots….as you have many times in the past.
One of the many reasons I did not subscribe to foreflight:
FlyQ puts ODP, ALT mins, DPs, STARs, and even the AFD all in the same area, indexed to the correct page.
Why not fly the instrument approach in reverse to this airport? That is published and easily available. I’m an ATP/CFII and have suggested this as a technique to IPC pilots. The FAA books as compared to JEPP are harder to follow. This accident reminds me of Reba McIntyre’s airplane departing Brown Field in San Diego to the northeast to stay out of the Class B airspace. With similar tragic results. Good article.
ODPs could certainly be written clearer. At my home airport (VGC), following the ODP text for RWY35 without also simultaneously meeting the implied climb gradient requirements (found separately) would result in flying into a windmill farm.
A charted ODP is good, and the FAA has started doing that–but VERY slowly.
But, be careful what you ask for. It’s a pretty day and you are wanting to fly IFR and just depart towards your destination. Do you want to be stuck following an IFR ODP by climbing over the airport? In non-radar, ATC will NOT be able to clear you direct.
IMHO, the best course is:
1. Chart ODPs
2. Include ODPs and departures in IPCs
Agree with John and all commenters. Perhaps another consideration would be to add ODP’s as a more specific training requirement under 61.65(c), which lists the required flight training areas of operation. It would then have to be flown and logged.
The crash that comes to mind is that of Frank Sinatra’s Lear carrying his mother. Departed out of Palm Springs on a dark night and turned directly into a mountain.
Thanks for referencing this accident. Actually, I think the problem was that they didn’t turn. Their clearance out of KPSP off of RWY 30 (now 31) was after departure, turn right direct TNP, maintain 10,000’. They missed the part about turning and stayed on runway heading at 10,000’. They hit near the top of 11,500’ Mt. San Gorgonio which is located northwest of KPSP. A contributing factor was that PSP approach was non-radar. The crew asked several times when they were going to be issued a turn and the controller thought they were referring to a turn at TNP. I lived near PSP in the 70’s and flew that departure often.
10 nm radius terrain elevation is given by ForeFlight when touching the mountain icon on the airport info sidebar that you get after touching on the airport on the map page. How about adding to this terrain info area a link to go directly to the Take Off Minimums for the airport for easier reference? Do this also in the Procedure/Departure area. I find it strange to have to go paging through a book of airports to find the ODP for an airport after already choosing the airport. I suppose this is due to the way the Take Off Minimums, ODPs, Radar Vectors info is presented by the FAA and creates some sort of conflict for ForeFlight.
Also make the mountain icon red.
Actually, I think the problem was that they didn’t turn. Their clearance out of KPSP off of RWY 30 (now 31) was after departure, turn right direct TNP, maintain 10,000’. They missed the part about turning and stayed on runway heading at 10,000’. They hit near the top of 11,500’ Mt. San Gorgonio which is located northwest of KPSP. A contributing factor was that PSP approach was non-radar. The crew asked several times when they were going to be issued a turn and the controller thought they were referring to a turn at TNP. I lived near PSP in the 70’s and flew that departure often.
John, you have brought up an issue that needs attention. Your very informative article captured the essence of the problem and lack of awareness of ODPs. At a mountain airport we frequent in Colorado, KPSO, there is an ODP, but I don’t think it’s often flown.
I considered how I would build and fly it with my GTN-650 and it took me some thinking and planning on my iPad with the GTN simulator to find a good solution. I’ve got the loaded as a route in my catalog now and if I ever need it I’ll have it readily available. But it was a little complicated to build and I’m not sure I would have done well had this been discovered in the airplane at the end of the runway. On a night or IMC departure from Pagosa Springs, with high mountains nearby, it’s serious business. I think we all need more awareness and training on ODP’s. Thanks for bringing it up.
Dan. ATP, CFII, 787 LCA.
All airports serviced by an IAP have an ODP. If there is no published ODP the following generic ODP applies: If no special takeoff minimums or Obstacle Departure Procedure (ODP) has been published, cross the departure end of the runway at least 35 feet high, and climb to an altitude of 400 feet above ground level at a rate of 200 feet per nautical mile or better before turning on course or to the heading assigned in our clearance.
While I can see the virtue in making this information more prominent, I have to wonder how often it would help in situations like this. I would never think to depart in the mountains at night without knowing the ODP or SID and thinking very carefully about how to get out safely (or not depart until morning). We can all make mistakes of course, but it seems like this pilot, RIP, just did not consider the situation carefully.
Sarah Fritts has an excellent article on this topic on her “Think Aviation” site.
Very sensitive and welcome subject! In the airlines there is no departure that not consider the obstacles alone – and we are talking about airplanes that can in fact climb with decent gradients even after an engine failure. So, when it comes to GA, where performance is often much more limited, it should be obvious to focus on that. The classic Bonanza crash in Telluride comes as a reminder, but other examples are not hard to find – imagine how many near misses we do not even get to know! Thanks for bringing this subject to light, John!
Excellent, John! As CFII for several decades, ODPs have seen to be a weak or non-existent point with transfer instrument students.
I agree, put ODPs out there on bold print! Thank You ☘️