Pelican's Perch #53:
Well, SIAP on You, Too!
There's a lot more to flying a Standard Instrument Approach Procedure (SIAP) than meets the eye. AVweb's John Deakin discusses what goes into the design of an SIAP and how to shoot an approach legally and safety — and tosses in some tricks, some gotchas, and some interesting case studies.
No, I'm not telling you where to go. Or perhaps I am, you figure it out.
"SIAP" is FAA-speak for "Standard Instrument Approach Procedure," and frankly, they're not very well understood. I don't profess to understand them fully, myself.
Yeah, you heard that right. I certainly don't begin to understand all the "stuff" that makes up one of these things. Does that sound strange, coming from a professional? Well, even the FAA's Chief Counsel gets it wrong now and then.
Now I ask you, if the big boys get it wrong now and then, what's a poor pilot to do? Well, one thing the pilot doesn't have to do is design the SIAP. Someone else has done that. The FAA has presumably surveyed the terrain rather carefully, established the minimum equipment to do it, and has set up the courses and altitudes to get there safely. The dirty details are a bit like watching what goes into making sausage, you really don't want to do it.
The pilot doesn't even need to know the criteria, because all that is built into the procedure on the Jeppesen or NOAA approach plate. The basic idea is quite simple, with only two rules in the language I speak:
- Don't hit nuthin'.
- Don't do nuthin' stupid.
Two more good rules (see, we're up to four, already!):
- What's legal may not be safe.
- What's safe may not be legal.
Our first objective in this Darwinian exercise is to get the airplane where it's supposed to go without hitting anything, to do that without scaring anyone (especially me), and finally, to satisfy the legal requirements.
The approach plate very nicely lays out the TRACKS you can use, and ALTITUDES you need to satisfy ALL the requirements, both for safety, and for the legalities. It even takes into account the SPEED of the airplane for such things as turns onto a new course, and of course, holding patterns and circling approaches. So, if you fly the TRACK depicted, and maintain at least the minimum ALTITUDES depicted, and fly the SPEEDS for your aircraft, that's about all you really need to know. The rest is just icing on the cake, academic in nature, and not as important as the "primary data." Jeppesen finally got the picture on this, and put the "primary" stuff across the top of the page, perfect for review and briefing other crew members.
There is only one exception to following the depicted approach tracks and altitudes, and that is if you're under radar control. Under radar, you're legal to do anything the radar controller says. So:
- Either follow the published SIAP, or follow radar instructions.
Now THERE is a rule I can live by!
Don't Hit Nothin'
If you follow the instructions from Approach Control, then it is up to the controller to place you ON the SIAP in such a way that you should be able to make the transition from radar to self-navigation with no unusual maneuvers.
That's not to say you shouldn't be aware of things to hit! If that information isn't in your own memory banks (from having flown there before), then you probably ought to review the situation from "all available information." That might even include having the appropriate Sectional chart available, even if it's not required under IFR. I consider those charts works of art, and so useful I'd rather have them than IFR charts, most of the time.
There are a couple of other issues I see coming up all the time.
- MSA (Minimum Safe Altitude). This is usually a little
compass rose at the top of the Jepp plate, with sectors, and minimum safe
altitudes in those sectors. THAT IS NOT OPERATIONAL DATA! It is there
simply to provide the pilot with some sort of hazy mental picture of where
the stuff is that you don't want to hit, and it may well NOT be
"safe" to go there. For one thing, it only provides "about
1,000 feet in mountainous areas," instead of the required 2,000 feet.
- MVA (Minimum Vectoring Altitude). Only the radar controller knows this one, we pilots just have to trust them. This can be the lowest of all the restrictive altitudes, and may be established to make it easier for the controller to vector you well below the MEA (Minimum Enroute Altitude), or below the altitude depicted on the approach plate. Do you trust the controller? Heck yes, I do, all the time. But it's also good to have a backup plan, in case your tire falls off and goes right through the radar antenna, knocking out the ATC radar for the day. Where would that leave you? In never-never land, that's where (and without a tire, too).
Case Study #1: Sunriver, Oregon
In a perfect world, we'd have radar vectors to final, then we'd have a "standard" approach, where the FAF (Final Approach Fix) is always at five miles from the threshold, at 1,500 feet, with an ILS to 200 feet, and a straight-ahead missed.
It ain't a perfect world.
How can we cheat, and remain legal? (Remember, if you don't know the rules, you can't cheat!)
Suppose you're shooting the VOR approach at Sun River, Oregon. This is an "interesting" approach that meets all the requirements, but just barely. The approach ends up in a bowl, with mountains all around, and will make you wish you'd done that overdue VOR check.
(In reality, the criteria call for the "protected airspace" to widen out dramatically, that far from the VOR, so "real safety" is not an issue. The limiting terrain at Sun River is the stuff just west of the long final.)
There is a 23-mile-long approach path from Deschutes VOR at 7,000', then stepdown fixes at 13 DME/7,000', 18 DME/6,000', and finally 23 DME at MDA. If I recall correctly (it's been a few years), radar will cover you to just beyond the VOR, then you're on your own.
Let's cover a few different scenarios at this airport.
Normal case, IFR. The controller will try to vector you to cross the VOR at 7,000, and clear you for the approach. Since it's a CENTER radar, not an approach control radar, such vectors aren't all that good (center radar has a much slower sweep, and the display is a mosaic from several radar sites), and they may not be able to get you down to that altitude. But there's plenty of room to get down from any reasonable altitude over the VOR.
What's a "reasonable altitude" in this case? There is an established gradient, but I don't know what it is, and I don't want to know, because I'll forget it, leaving those brain cells dead forever. I've got a lot of dead brain cells from trying to remember stuff I didn't need (and a few others, too).
But figure it out for yourself. Do a rough WAG of your groundspeed (or true airspeed) in miles per minute (round it off, we're not working with interstellar trajectories, here), how long it will take you to travel that distance (nearest minute), and how fast you have to descend. Do it, and get on with the show.
Note that if there was a procedure turn depicted, and radar wasn't rock solid, you'd have to do the full procedure, hit the VOR, turn outbound, do a procedure turn, and THEN start inbound. "Do what's written, or use radar," remember?
You should maintain 7,000' to 13 DME, then descend to cross the 18 DME fix at 6,000'. I don't care very much how you do this one, you can "dive and drive," or you can time it against a rate of descent, but I'd sure like to be AT 6,000' when I hit that 18 DME fix, to get a good start on the next stepdown, a far more important one. At that fix, I'll "dump it to the max," trying for 1,000 fpm or better. I want to GET DOWN to MDA, level out there, and start looking as soon as I can. For a couple of previous columns on this, see the following links:
If you don't see anything at 23 DME, it's a miss, of course. The interesting thing is doing this in good weather once, so you can see the terrain. When you hit that missed approach point, you are probably at 5,480 feet, and there is a 5,800-foot mountain at your 9 o'clock, roughly five miles. The pullup is a left turn back to the VOR right toward that mountain. It looks awfully big, and awfully close in your windshield as you go through the ninety-degree point in the turn! Of course, if you're IFR, you can't see it, which is much better.
(Again, the reality is that there is plenty of terrain clearance, this missed approach isn't even "close.")
Note the Maltese Cross at 18 DME, making that the FAF (Final Approach Fix). If the controller could vector you that low (unlikely), he could give you vectors to any intercept before that, and turn you loose. But from memory, I don't think they have coverage that low, so you're stuck with the full approach, from the VOR.
Now, suppose there is a monster thunderstorm over the VOR, and a widespread solid layer of clouds everywhere else, including over the airport, with tops at, say, 5,000 feet.
You're arriving IFR in your Bonanza (Part 91), you don't want to go through the boomer, and radar says he can't vector you for whatever reason. With a little coordination, a little timing, and a cooperative controller, you could either request a VFR restriction, or even cancel IFR, position yourself anywhere on the approach track while maintaining VFR/VMC (in this case, "VMC" means "Visual Meteorological Conditions"), and then call, "Established on the approach, request IFR clearance for the approach." Make sure you're not interfering with anyone else while doing your maneuvering, of course, and remember minimum vertical separation from the cloud layer. This is ONLY a good procedure if you're well in the clear, with no restrictions to visibility. If that's not true, you'd better be intimately familiar with the area!
Case Study #2: Oxnard, California
I use a variant of this trick at Oxnard, Calif., for training, all the time. Pt. Mugu Approach Control has good radar coverage over the area, and OXR has a full ILS, but the missed is almost straight ahead to 4,000', then there is a LONG downwind vector, often ten miles or more to set up for the next approach. We might get two approaches in an hour, and the long vectors are agonizing. I'd like to improve the efficiency (read: cost) of training.
There is often a "Marine Layer" present, bringing ceilings and visibilities down to minimums, with tops usually around 1,200 feet. WONDERFUL training weather (but not for CAF aircraft, limited to 1000 & 3). I love those days, with my own airplane! On the miss, I'll climb to 1,000 feet above the tops (remember the cloud clearance rules?), and cancel IFR. The instant Mugu acknowledges that, I'll do a steep turn back toward the FAF, wide open power, do an 80/270 course reversal, again with a steep turn (good practice, and it helps to kill the speed off), roll out on final on the LOC and GS, and ask for an IFR approach clearance. Do the approach at 160 knots, pull up and repeat. I think I got 10 ILSs done in about an hour, one day. That sure will bring the old scan back up!
Mugu is military, very conservative, and the first few times the controller will be expressing the attitude of the baby eyeing a large rectal thermometer, "You want to do WHAT?" Generally, by the third approach, he gets the idea, and things go pretty smoothly after that.
Yet another variant of this is instrument training in good weather, same airport. Asking for advisories from Mugu is the kiss of death, because they absolutely require vectors, and will not allow "self-navigation." The old airplanes often don't climb very fast, so you can easily be 10 or 15 miles out to sea on the missed approach before turning back towards the downwind vector, then they want to vector us at least 10 miles east, a long base, and turn final 10 miles out. That can quickly turn into a 50-mile pattern, unacceptable to a charitable organization, flying an airplane burning $500 per hour in fuel!
Instead, we'll just stay on the Oxnard tower frequency, and do our own self-vectors, climbing to 2,100' to stay just above the Camarillo Class D, which is roughly under the ILS FAF for Oxnard. That allows an immediate right turn right after the missed (or the takeoff), a ten-mile downwind, a fairly tight turn to an intercept, and "Cleared for the simulated approach." Half the flying distance, twice the approaches. Meanwhile, as far as the tower is concerned, we're just doing a large VFR pattern, reporting, "Ten east, request the straight in," with no mention of any instrument approach. If we mention the word "ILS," they will instantly say, "Contact Mugu Approach," which is just what we DON'T want to do. This also rather neatly avoids noise-sensitive areas, the Mugu ILS final, and the VFR pattern for CMA, too. Both the OXR and CMA towers are absolutely excellent at coordinating all this, even to the point of checking with Mugu for inbound traffic! Both towers have BRITE scopes as well, and often call traffic for us. It really works smoothly (at least from my viewpoint!) Most of the controllers seem to be glad for the novelty of having a dinosaur in the pattern, and some of the things they get to see are amusing, to say the least!
But, back to our approaches.
The rationale is that pilots simply don't have the information to determine when it's safe to do anything not depicted, without radar. The criteria are so complex that there is simply no way we can do it.
Case Study #3: Salina, Kansas
Take the Salina, Kansas ILS 35 approach, for example. For the arc procedure, you've got to have radar, OR you must start the arc from the SLN 260/23.0 or the SLN 075/23.0. You cannot, without radar, start the arc anywhere else, because that hasn't been surveyed fully. For example, while it may meet the requirements to be ON the arc (from the IAF), it may not satisfy all the requirements for a TURN onto the arc anywhere else. That arc is NOT a "family of IAFs" — there are only two. Sure, there may be no problem here, but how about all the rest of them in the USA? Additional IAFs COULD be established, but that would add a lot of work and flight inspection. "Personal Knowledge" is no part of the procedures, and simply doesn't count.
The language about starting only from an IAF (without radar) was added in 1994, but the intent was always the same. Too many people were short-cutting things, overshooting the turn to the arc, or setting themselves up too high, so the FAA added the verbiage.
This business of "dropping in on" an arc, or any other point on any approach without radar is illegal when IFR. If you want to "cheat" using the VFR trick, be aware it may have risks (even though legal), and is valid only for a pure Part 91 operation, for it will not be approved in any "OpSpecs." Remember, Part 91 is "permissive," in that anything not mentioned is permitted, but the other regs are "restrictive," in that nothing is permitted, unless specified.
My thanks to Wally Roberts for his helpful comments in preparing this column. He is the absolute world authority on these issues. I think he is perhaps the only person alive who truly and fully understands them, for he has made this area his life's work. He also designed the approach at Sun River, years ago. For a special treat, check out Wally's TERPS Web site.
Be careful up there!