Sunday, January 28, 2007

Flying RNAV Departures

My apologies for the long break from blogging. This was due mostly to my getting sick and then having a spate of unplanned maintenance activities on my Jetta TDI. We're both now in fine fettle, thank you, though I'm a bit poorer than I'd like. These recent events have led me to offer some unsolicited testimonials:

Bauer's European Auto Repair in Berkeley - they actually understand diesel engines, they're friendly, and their prices are reasonable, and they don't treat you as if you're in league with the devil if you are using biodiesel.

Zicam Cold Remedy Rapid Melts - a homeopathic cold remedy recommended to me. I was skeptical until I tried it. While it doesn't make your cold go away, it turns the volume down on the symptoms and helps you get on with your life.

Mucinex - a time-release, tablet version of guaifenesin (a common active ingredient in cough syrups). It's better than cough syrup by a long shot.

Now on to the promised discussion of RNAV departure procedures. RNAV departures in the U.S. are designed primarily to simplify the delivery of IFR clearances in busy airspace, to provide obstacle clearance in remote areas where VOR reception is spotty due to terrain, or both. They come in a couple of flavors: Type A and Type B. Both types usually start with a heading or vector from the end of the departure runway to an RNAV fix. The differences between type A and B SIDs are largely incidental to the pilot flying with a 430W or 530W, but here they are for completeness:

First RNAV fix from the end of the departure runway:
Type A: approximately 10NM
Type B: "near" the departure end

RNAV accuracy required during 95% of the flight time:
Type A: +/-2NM
Type B: +/-1NM

Altitude Engagement:
Type A: no later than 2000 feet Height Above Airport
Type B: no later than 200 feet HAA

Equipment required (described in AC 90-100)

There's really not a lot to flying a RNAV SID, as long as you do some preflight planning. The first task is to ensure that you can meet any required climb gradient specified on the chart. Consider the CHOIR ONE departure from Alturas, California.

And I shouldn't have to say this, but:

Don't use any of the charts, snippets, or other illustrations shown in this blog for navigation in a real aircraft.

This SID requires a minimum climb of 367 feet per nautical mile or 389 feet per nautical mile, depending on the departure runway, up to 11,000 feet. The climb gradient is expressed in feet per nautical mile because the procedure designers don't know how fast your particular aircraft is traveling in a climb. Here's a snippet from an old Klamath Falls VFR sectional showing the topography around Alturas.

Jepp charts are kind enough to calculate the required climb rate for you, translating the climb gradient into feet per minute for a range of different climb speeds. If you use FAA charts, you can use a table in the front part of the approach booklet to do the conversion or just use your calculator:

(398 feet/NM X 100 NM/hour)/60 minutes/hour = 648 feet/minute

This climb rate at altitudes above 6,000 feet MSL is beyond the capability of many single-engine GA aircraft, so check your aircraft's flight manual. If you're flying a multi-engine aircraft, of course you'll want to look at the single-engine rate of climb performance, just in case.

Next, you'll want to do either an FDE prediction and/or a RAIM prediction prior to departure. Power up you Garmin GPS, bring up the Auxiliary pages, find RAIM prediction, and tell the unit to do it's thing. Remember that RAIM prediction uses a satellite almanac to predict the number of satellites your GPS antenna should be able to see, but it is not a guarantee that the in-flight RAIM calculations that the receiver makes will always succeed. Translation? Your RAIM prediction could succeed, but you could still get a RAIM failure in flight (though it is unlikely).

Since a pre-flight RAIM prediction is so important, why does Garmin insist on burying this feature, requiring a bunch of knob twisting and button pushing?

For this example, we'll do a trip from Alturas to Reno, Nevada using the CHOIR ONE departure, then direct to a fix on the RNO RNAV 16R approach. Press the FPL button, press the small knob on the lower right to enter cursor mode, scroll to the empty space below KAAT, then turn the small knob to spell out KRNO for Reno.

Press the PROC button, then scroll with the big knob on the lower right, highlight SELECT DEPARTURE, and press ENT. Select CHOIR1 from the list, and press ENT.

On departure, you can use the terrain view to give you an idea of where the obstructions are located as you climb.

As you approach the second waypoint, the 530W will give you a 10 second countdown before you need to change heading.

As you approach CHOIR, you'll get the same 10 second countdown, then the unit will tell you to turn. As I've said before, this countdown and turn instruction are a huge improvement from the earlier 530/430 units where the prompt was so short lived that it was easy to miss. Turn when the unit tells you to turn and you'll fly by the waypoint and get established on the airway without overshooting.

In a future installment, I'll cover flying a STAR.

Thursday, January 18, 2007

Different is not the Same

After writing about the changes to the OAK ILS RWY 27R, an approach I fly frequently, I started to apply my Jepp updates this morning. After getting all the new charts inserted in the binder, I turned to retrieve the new ILS 27R chart. I like to keep frequently used charts in plastic protectors and store them in my kneeboard. So imagine my surprise when I found this chart had the same date and version as the old one.

With very little time before I had to be on the road to the airport, I visited Jeppesen's web site to get a customer support phone number. They don't make it easy, but eventually I found the correct number. I dialed it and sat on hold for a few minutes. Then I decided to see if there was an email address. I finally found it and sent a message asking why this particular approach plate seemed to be the old version. After teaching a G1000 checkout flight, I came back to find an email message from Jepp.
Hello John,

The NACO chart is incorrect due to the procedure changes being postponed by the FAA and NACO not receiving notification of the postponement.

An FDC NOTAM is being written at this moment. The KOAK ILS or LOC/DME Rwy 27R is NA at this time.

I thank you for writing as your attention to this lead us to investigate and as a result the FAA is responding quickly.

We also thank you for using our products.

And over my morning espresso, I read a Reuters article entitled Boeing says new 747 interior frees travelers to fly that included this statement:
"We discovered that today people in many cases just like to numb themselves to the experience (of flying)," said Klaus Brauer, Boeing's director of passenger satisfaction and revenue at a demonstration of the plane's new interior.

"We are trying to get deeper than things people can articulate."

Whatever that means ...

Lastly, I finally got time to view the old, government training film on flying the F4U Corsair that I had first read about over at Land and Hold Short. Check it out!

Wednesday, January 17, 2007

Changing and Adapting

A former instrument student of mine sent me email a while back with a question about the Salinas ILS RWY 31 approach. His question had to do with the missed approach and he wondered why the approach chart didn't list ADF or DME as being required. I looked at the approach and I, too, was puzzled.

The missed approach holding fix is MARNA, defined by 10.9 DME fix on the SNS 275˚ radial or the intersection of the SNS 275˚ radial and the 207˚ bearing to the MUNSO locator outer marker. Without ADF, DME, or GPS there would be no way for a pilot to identify the missed approach holding fix. So I suggested he email the FAA and report it as a potential error. He did and got a fairly prompt response, thanking him for pointing out the error and assurances it would be fixed in the next revision. Within a week a NOTAM (notice to airmen) appeared saying ADF or DME was required for this approach and, true to their word, the next version of approach charts showed the change. "The system works!" my former student observed.

I regularly use FAA and Jeppesen approach charts, since I may give instruction to pilots who prefer one or the other. I include the FAA charts in examples on my postings here because they are easy to acquire and upload. While looking at the OAK RNAV (GPS) 27L approach, which I used in a recent post, I noticed an error.

The inset map shows the approach course not being aligned with the runway centerline. The course depicted seemed to be for the OAK VOR/DME 27L approach (the Oakland VORTAC is slightly south of the runway centerline). Since the RNAV 27L approach course should be aligned with the runway, I followed my own advice and sent the FAA an email. I also received a prompt reply, thanking me for pointing out the error and an assurance that it would be fixed in the next revision. This got me to thinking about another approach conundrum - the OAK ILS 27R.

The Oakland ILS approach to 27R was changed about four years ago and though I no longer have a PDF that old version, I remember it well. The final approach fix was CASES, 5 NM from the threshold, and it had an outer marker. Four years ago, for reasons that no one seems to know, the FAF was moved out to 5.5 NM. it was named FITKI, and the glideslope angle was reduced slightly. The glideslope intercept was moved in 3.5 miles from URZAF (9 miles out) to FITKI. The odd thing was that the outer marker was still there, it was depicted on the approach chart, but it was no longer part of the ILS approach! I continued to fly with pilots who performed this approach and started their timing to the FAF when they saw the outer marker light flash on their marker beacon receiver, blissfully unaware that they had already passed FAF. I also flew with pilots who still continued to intercept the glideslope at URZAF instead of stepping down to 1500 feet to intercept the glideslope.

Yet after all these changes, the CASES outer marker still had an approach function, at least until last year: It was a step-down fix for the OAK NDB RWY 27R. Alas, the NDB approach is no more. And last year, the ILS approach name was changed to OAK ILS or LOC/DME 27R because it became clear that, without a marker beacon to designate the final approach fix, you'd need DME to fly the localizer-only approach else you wouldn't know when to start descending.

Last week I started nosing around the NACO site and discovered this page where I learned that the OAK ILS RWY 27R will soon be changing again. How soon? Well today!

The final approach fix has been renamed from FITKI to CUVSA, URZAF has been renamed to UPACI, and the missed approach holding fix has been move from PEERE intersection all the way out to REBAS intersection. It appears the old CASES outer marker may finally have been decommissioned and the RORAY locator middle marker has also gone away.

I'm not sure why some of the changes are being made, but one thing is for sure: All of the pilots, instructors, and air traffic controllers will have some adapting to do.

Monday, January 08, 2007

Flying LPV Approaches

NOTE: For a more up-to-date discussion, see Understanding RNAV Approaches.

Continuing my series on the simulating RNAV approaches with the new Garmin 400/500W Trainer software, here are some preliminary results of experiments flying an LPV approach.

I chose the KOAK RNAV 27L approach, which has published LNAV, LNAV/VNAV, and LPV minima.
I initialized my position over TRACY intersection (northeast of the Livermore airport) at 6000 feet, clicked the direct button, and selected KOAK as my destination. Clicking the PROC button, I scrolled down and selected RNAV 27L from the list of choices, and selected SUNOL as the initial approach fix. Since you often get vectored to a fix inside the IAF on approaches into Oakland, I clicked the FLP (flight plan) button, clicked on the small knob on the lower right to enter cursor mode, scrolled to the HEDER waypoint with the big knob, and clicked on the direct button. By the way, loading an approach procedure and then proceeding directly to a fix on that approach effectively activates the approach.

Approaching HEDER, I used the Trainer's throttle slider switch to reduce my airspeed to about 120 knots and the countdown message appears, providing a 10 second warning for the turn to join the intermediate approach course. It's harder to see in this example, but there is a small arc showing the path to fly-by HEDER and become established on the approach. Note that the course sensitivity annunciation is still TERM (terminal mode), meaning a full scale deflection represents a course deviation of 1 nautical mile. Once past HEDER and established, it will be time to descend to cross VODSY at or above 1600 feet. Notice that there is still no clue as to whether the unit will provide LPV, LNAV/VNAV or plain old LNAV sensitivity, so it's best to be prepared for the worst - a non-precision descent to 420 feet and a minimum 5000 foot RVR. Just in case there is going to be vertical descent guidance, it would be best to be at 1600 feet a mile or so before VODSY.

After passing HEDER and joining the approach course, the course sensitivity annunciation changes to LPV, meaning the 530W has the required GPS accuracy to provide vertical guidance and I can use the lowest (LPV) minima shown on the approach chart - 259 feet and a 4000 foot RVR (runway visual range). I just hit the jackpot - the lowest minima - though there won't be a flood of nickels pouring out of the 530W. Descending on the glideslope with the trainer is a bit tricky since you aren't really flying a plane, you're just clicking on an altitude button to descend or climb 100 feet at a time. Descend too quickly when using a 530W that has TAWS and you'll get a terrain warning, even with the Trainer software.

The next installment will investigate flying RNAV departure procedures with the 400/500W Trainer.

Friday, January 05, 2007

Flying LNAV Approaches

NOTE: For a more up-to-date discussion, see Understanding RNAV Approaches.

A while back, I took Garmin to task because it seemed their new WAAS-certified GPS units promised to increase pilot workload at a critical moment in flight - just before the final approach fix on an RNAV approach (like they listen to me!). Well, it might not be as bad as I first thought because these new units also provide some new features that could really help a pilot stay ahead of the game. Continuing where I left off with my previous post on FDE prediction, here's an example of flying the Byron RNAV RWY 30 approach using the new 400/500W Trainer software.

There are several new features in the 400/500 WAAS-certified unit, many of which are described in this Garmin document. Some of the features I find most useful don't seem to be mentioned.

One change involves the GPS status screen. Note that there are some new terms used, namely HFOM and VFOM. These stand for Horizontal Figure Of Merit and Vertical Figure Of Merit, which represent the 95% confidence in the horizontal and vertical accuracy. The Garmin 530W Pilot Handbook and Reference simply says that smaller numbers mean greater accuracy. EPU (Estimated Position Uncertainty) is more straightforward - it's the estimated horizontal position error in nautical miles. Lastly, the histogram bars that show the relative signal strength of each satellite will contain a D when that particular satellite has differential corrections (WAAS) applied. This is good stuff to know, but it only indirectly helps you fly an approach.

One annoyance is that when using the Trainer, the MSG annunciation is constantly displayed and when you press the MSG button you're warned to not use the simulator for navigation. This is important if you are using an actual 530W unit in simulator mode, but for the trainer software to display this message continuously it is pointless and annoying.

Back to the Byron RNAV RWY 30 approach, let's say you've loaded the approach and are being vectored to SHARR intersection (the initial approach fix). As you approach SHARR (a fly-by waypoint), you'll begin to see a curved path displayed which shows the path you'll fly if you follow the turn anticipation prompt provided by the 530W. I find this to be a huge improvement because it gives the pilot a visual cue as to the path they'll be flying and helps them visualize and prepare for the change in course.

The prompt for turn anticipation messages in the old 530 units lasted only a few seconds and unless you watched the display like a hawk, they was easy to miss. A new feature in the 430/530W is a 10 second countdown display prior to beginning a turn and prior to any waypoint passage. This message is much harder to miss and the countdown helps get the pilot in step with the approach.

A few miles outside of the final approach fix, note that the course sensitivity annunciation has changed from TERM to LNAV/+V. Now as I understand it, this annunciation is for approaches that have published LNAV/VNAV minima, but haven't yet been identified as such in the Jeppesen database that Garmin uses. The Byron RNAV RWY 30 approach has only LNAV and LPV minima, so I'm concluding that LNAV/+V is also used to provide advisory vertical guidance on approaches that have only LNAV minima. So LPV, LNAV/VNAV, and some LNAV approaches will all display vertical guidance and it's up to the pilot to note the annunciation (LPV, L/VNAV, or LNAV/+V) and descend to the appropriate minima. If the LNAV annunciation appears, they you won't receive any vertical guidance.

When vertical guidance is provided, the behavior similar to an ILS glideslope and you should intercept the glideslope from below, just like you would an ILS glideslope. You then descend using that guidance to the appropriate MDA or decision height. It appears from my use of the simulator that this vertical guidance is provided right down to the runway, so you'll need to know when to call it quits and level off if you don't have the required visual references to operate below the MDA or DH.

You can use the altitude button on the Trainer's simulated autopilot to descend on the advisory glideslope, keeping the needle centered. If you descend at too high a rate and you have selected to simulate a GPS unit with TAWS, you'll even get a terrain warning. Should you need to execute the missed approach, the unit's behavior is the same as in the old 430 or 530 units.

Next installment? LPV approaches with the 530W.

Tuesday, January 02, 2007

Finding (and Excluding) Faults

A reader recently did me a big favor and pointed me to Garmin's newly released PC-based 400/500W GPS Trainer software. As soon as I found time, I did the sensible thing and downloaded the new Trainer and started playing with it. You should, too. Here's why.

This free Trainer software allows you to simulate any of the new WAAS-certified 400 or 500 series GPS units which are approved as a sole means of navigation. You can also select to simulate terrain awareness (TAWS). The sim has a relatively new GPS database and will let you experiment with flying RNAV approaches. The trainer software also plays an important role for pilots who purchase or use the actual WAAS-enabled 400/500 GPS units because it provides an FDE application, which stands for Fault Detection and Exclusion. FDE will become an important part of the preflight planning chores for pilots using one of these units in a real aircraft. More on FDE after reviewing some few pointers on setting up the Trainer.

And let me say "Kudos!" to the folks at Garmin for providing this Trainer free of charge. This is some of the best marketing imaginable, especially since the Trainer is infinitely more useful and usable that their G1000 trainer.

When you first start the trainer, you'll see a window like the one shown at the top of this blog entry. Use the Options pull-down menu, select Unit, and then choose the type of GPS you wish to simulate. I chose the 500 series with TAWS because the 500 has a larger screen and can display more data fields on the two moving map views it provides. Power up the unit by clicking on the communications volume knob. You'll see the unit comes alive thinking you are somewhere in Kansas. I selected the Options->Initialize Position menu item and used the large and small GPS knobs on the lower right side to enter KOAK (Metropolitan Oakland International Airport) as my position, since I live, work, and fly in Northern California

You may recall me complaining that it seemed that pilots flying RNAV approaches with one of these WAAS-certified units would need to be prepared to fly the approach to two or three different minima, based on the GPS accuracy available at the time the approach is commenced. Well the FDE application takes some of the sting out of this by providing a prediction of the sort of GPS course accuracy you can expect, based on the date and time of your flight, your departure, your route, your destination, and your phase of flight. In fact, if you plan to use one of these WAAS-certified receivers as your sole means of navigation (oceanic flight, more that 200 nautical miles from any airport), you must use the FDE prediction application prior to beginning your flight.

But even those of us flying RNAV SIDs, STARs, and approaches should consider making FDE prediction part of our pre-flight planning when possible. More on the implications of this requirement in a bit. At the very least, make sure you obtain GPS NOTAMs during your pre-flight briefing, such as choosing the DUAT option shown here.

To use the Trainer's FDE prediction application, you first need to enter a flight plan into the unit. So I created a simple flight plan from Oakland to the Byron Airport, about 30 nautical miles to the east. I chose Byron because they have a recently updated RNAV approach with published LPV minima. After entering the flight plan, I pressed the PROC button and loaded the RNAV RWY 30 approach with SHARR intersection as the IAF. In real life, it is unlikely that ATC would let you proceed to directly to SHARR as a transition when approaching from the west, but I was just simulating.

With a flight plan entered and approach loaded (but not activated), you can launch the FDE program using the Options->Fde Prediction Program menu item. For me, this resulted in a series of dialogs telling me that the GEO almanac and the configuration file were corrupt or out of date. Each dialog gave me the option of downloading the current almanac and configuration file, so one assumes, and this is important, that a valid internet connection is required to use the trainer's FDE features.

Finally, the WFDE prediction program was launched providing more options. One option allows you to select the GPS antenna in your aircraft. I actually don't know what sort of antenna might be installed on the average GA aircraft, so I just left it on the default A-33 model. You can also specify your departure date and time. Finally there is a Setup button, which you should click on before clicking on the Calculate button.

The Setup option lets you specify some important facts concerning your flight, like the estimated ground speed, the type of route, such as Oceanic, EnRoute/Terminal, etc.

Now click on Okay, then Calculate, and the FDE application comes back with a display telling you if GPS is available for your route and whether RAIM and FDE are available. The program also tells you which RNAV approach minima you can use for planning purposes, though this is not a guarantee that the lowest minima will be available when you actually arrive.

So what's the big deal about FDE, you ask? Here's some guidance from the Garmin 530W Pilot Guide and Reference (with hyperlink added by me):
The Garmin 500W-series software incorporates a Fault Detection and Exclusion (FDE) algorithm, thus providing a basis for approval per FAA Notice N8110.60 requirements for "GPS as a Primary Means of Navigation for Oceanic/Remote Operations."

The FDE consists of two distinct features, fault detection and fault exclusion. The fault detection feature detects the presence of an unacceptably large pseudorange error for a satellite (and presumable, position error) for a given mode of flight. Upon detection, fault exclusion follows and excludes the source of the unacceptably large pseudorange error, thereby allowing navigation to return to normal performance without an interruption in service. To enhance safety, FDE functionality is provided for other phases of flight (non-precision approach, terminal, en route). The FDE functionality for non-oceanic flight phases complies with missed alert probability, false alert probability, and failed exclusion probability specified by DO-229C/TSO-146a.
The WFDE Prediction Program Instructions go on to clarify what an operator (or pilot) must do if they plan to use the unit to fly RNAV routes, departures or approaches (again, hyperlinks added by me).
Prior to navigation on U.S. RNAV routes. SIDs, and STARs (reference AC 91-100) all operation of Garmin WAAS-certified units must utilize the WFDE Prediction Program to determine RAIM availability.

All operators of Garmin WAAS-certified units should utilize the WFDE Prediction Program to determine WAAS satellite visibility when planning an LNAV/VNAV or LPV approach.
To my mind, this means that prior to departure you'll want to have access to a computer that has the Trainer software loaded and that computer will need to have an internet connection. Ideally, you'll want run the FDE prediction application before you fly if you plan to use a 400W/500W series GPS for LNAV/VNAV or LPV approaches, though this might not always be practical.

In the next installment, I'll detail some simulated RNAV approach adventures using the Trainer software.