Probe Into Fatal UPS Caravan Crash Presents More Questions Than Answers


Investigators have no shortage of possibilities to explore in the April 13 fatal crash of a Cessna Caravan flying a regular package-delivery route under contract to UPS. The 30-year-old pilot, Brittney Infanger, was killed in the crash after attempting her second non-precision GPS instrument approach in misty, snowy weather to Runway 20 at city-owned Burley Municipal Airport (KBYI) in Idaho. The approaches followed a normal one-hour flight from Salt Lake City International Airport (KSLC).

The Caravan’s right wing apparently struck a chimney stack on the roof of a potato-processing plant located under the approach path less than a half mile from the runway in the town of Heyburn, Idaho, just across the Snake River from the airport.

Infanger’s father, a pilot, has publicly called on the county to close BYI as unsafe, and replace it with a new airport in a less hazardous location. He claims, along with other pilots in the area, that industrial growth surrounding the airport, the potato plant in particular, has compromised the approach paths. “They’ve allowed this potato processing plant to continue to expand,” he told a local news outlet, “and this chimney comes up and has a huge amount of steam. If the wind is blowing [a certain direction], you fly right into this wall of steam. That was the case that morning.”

According to YouTuber Juan Brown on his accident-analysis channel, the RNAV (GPS) 20 approach to BYI does meet FAA’s Standard for Terminal Instrument Procedures (TERPS) requirements, though the step-down segment of the approach calls for a steep (3.7 degrees rather than the standard 3 degrees), unstabilized close-in descent phase after passing the chimney stacks.

The accident flight left SLC at 7 a.m. local time and was vectored for the RNAV approach to Runway 20. Infanger executed a missed approach on the first attempt and was vectored through the missed approach procedure for a second try. ADS-B data—which can be inaccurate—show her descending below the minimum descent altitude before hitting the stack and crashing inverted onto the plant’s rooftop. No one on the ground was injured.

Weather at the time was reported with winds from the south at 8 knots, visibility at one statute mile in light snow and mist, a broken ceiling at 800 feet and overcast at 2,300 feet. The temperature/dewpoint spread was -3/-5 degrees Centigrade.

Readers of the Kathryn’s Report summary of the accident—some of whom identified themselves as local pilots—have suggested numerous possibilities for the premature descent below minimums, including fuel starvation; icing; and that, with limited visibility and high pilot workload, Infanger might have mistaken the rooftop of the potato plant for the runway, which has only runway end identifier lights (REIL) and no visual approach slope indicator (VASI). The Caravan’s Garmin G1000 flight deck could provide investigators with stored data that would be more accurate than the available ADS-B readouts.

Brown reported that Infanger, a native of Salmon, Idaho, completed training with American Flyers and received her instrument, commercial and flight instructor ratings in May 2019. Described as a skilled and experienced pilot, she was reportedly flying Caravans for Gem Air under contract to UPS to build hours toward her airline transport pilot (ATP) rating for an airline career.

Mark Phelps is a senior editor at AVweb. He is an instrument rated private pilot and former owner of a Grumman American AA1B and a V-tail Bonanza.

Other AVwebflash Articles


  1. It would not be the first time an approach had something appear into the approach path (trees or man made objects) that was not there when the approach was first implemented. The most obvious mistake made is that the pilot tried another approach after going missed the first attempt. Should have diverted to alternate. As far as whether she went below MDA, or anything else happened, the investigation will attempt to find that out. I used to fly Caravans doing the same type of flying she was. That approach is a little more challenging than ones flown in flat land. I have seen reports of other pilots in the past ducking a little lower with fatal results. I have even substituted for another pilot on a run who died doing exactly that. Not trying to make accusations in this accident. As I started to say, could very well be an object was built into the approach descent path that was not there when approach was designed. RIP.

    • How do you hit a 60 foot high object with a 409 AGL MDA without going below the MDA? I ran the numbers, the top of the chimney is at a 1.37 degree path from the end of the displaced threshold. Even more shallow from the touchdown zone. The VDA is 3.75 degrees with a threshold crossing height of 40 feet. Think about that last number.

  2. “GetThere-itis”? The pressure to make a cargo stop (and the repercussions for not doing so, when others have) are probably quite high.

    I’m reminded of another similar story years ago of a young (woman) pilot who died flying a package delivery Caravan into a TRW somewhere in California.

    “… there are no Old, Bold pilots.”

  3. From the FAA’s Part 77 Obstruction Evaluation (Aeronautical Study No. 2016-ANM-1617-OE)

    Agglomerate Exhaust Stack

    It is determined that the proposed structures would have a substantial adverse effect on the safe and efficient use of navigable airspace by aircraft at the current airport location. However, planning is currently underway to identify sites for a potential airport relocation. The airport relocation planning may or may not be successful in moving the airport.

  4. It would not be the first time a building project, whether a cell tower or a steam-stack had been built into an approach path and exceeded the blueprint specs. If the FAA had not yet tested the approach path, or missed the obstruction due to flying the approach at night (I’ve been part of such a sequence, and sometimes it takes months and even years to conduct the test), the threat goes from theoretically safe to fatal.
    Also, the difference in a 3.0 deg. and 3.7 deg glide path may not sound like much, but depending on the wind, whereas a 3.0 deg will average ~7 – 800fpm, 3.7 can be as much as 11 – 1200fpm. For those who have depended on steep NDB approaches to get back to base, the difference can be dramatic in low weather.

    • To be at the top of the chimney (not low enough to catch a wing), she was on a 1.37 degree path to the end of the displaced threshold. A BIT shallow. And low for that far out. Are you at 60 feet AGL at 2500 feet from the end of the runway usually?

  5. My condolences to the Infanger family for their tragic loss.

    Where was the Director of Operations, the flight following, in the hours before this fatality? Where were the checks and balances in their FAR135 operations? The spud factory wasn’t a surprise to anyone. Everybody knew the factory was there. So……..what special procedures did the 135 operator have for what appears to be an extreme condition? None likely.

    The DO for this 135 needs to be held fully accountable for this totally preventable tragedy.

    God bless.

    • Without full understanding of the facts (unless speculating), it is not fair to hold the DO accountable nor hypothetically a lack of “checks and balances”. Maybe there were hundreds or thousands of flights into KBIY without incident. Remember, this is a Part 135 flight (I presume) not a Part 121 that requires an FAA certificated flight dispatcher and signatures from both PIC and dispatcher.

      Unless there was a mechanical problem with the Caravan, it must have been too low over the plant. It appears that the stack is at 4304 ft MSL, below the MDA of 4560 ft. Aggravating factor could have been effective visibility less than 1 sm and lack of a VASI. (Note there is a VASI on RWY 02.) How would “checks and balances” have made a difference?

      • The entire reason for a DO is to provide a layer of qualified leadership, responsibility, control, beyond that of just the pilot. Whether you agree or not the DO had operational control of this flight. Fuel, icing conditions, low visibility, a known hazard, a steep approach, pilot qualifications, required rest period, are all responsibilities of the DO as well as of the captains.

        Just because the approach was legal in those weather conditions doesn’t automatically make it safe. It is the DO’s responsibility to see pitfalls before they become problems. The margins for error were slim to none on this approach. What if the steam from the spud factory flash-iced the windscreen (and de-ice plate) just long enough to distract the pilot during the visual transition phase?

    • Jeff, why are you so eager to hang the DO based on your own (incorrect) presumptions? One thing we CAN assume is the pilot clearly went below minimums…where is your criticism of that performance that directly resulted in the crash?

      Basic airmanship (e.g. staying at MDA) would have prevented this “totally preventable tragedy.” So the PIC, not the DO needs to be “held fully accountable.”

      You’re welcome.

      • The entire reason for a DO is to provide a layer of qualified leadership, responsibility, control, beyond that of just the pilot. Whether you agree or not the DO had operational control of this flight. Fuel, icing conditions, low visibility, a known hazard, a steep approach, pilot qualifications, required rest period, are all responsibilities of the DO as well as of the captains.

        Just because the approach was legal in those weather conditions doesn’t automatically make it safe. It is the DO’s responsibility to see pitfalls before they become problems. The margins for error were slim to none on this approach. What if the steam from the spud factory flash-iced the windscreen (and de-ice plate) just long enough to distract the pilot during the visual transition phase?

  6. Presumably not a new airport on a feeder route. So why is the potato plant an issue now? Unless the tower penetrates etc.

    So low below MDA has nothing to do with steam. It’s situational awareness that was lacking, or it was the urge to duck and see was a great pull for her.

    Amazing that no post-crash fire occurred either.

    Either way I’m sorry for the famiiy’s loss may she Rest In Peace and may the family and friends regain a sense of normalcy and closure in the coming months and years.

  7. I took a TERPS class years ago as an FAAer. I was totally surprised to find that the FAA has no authority to stop construction projects in proximity to an airport. The construction project proponent does have a requirement to advise the FAA of their proposed project. And then all the FAA can do is evaluate it by TERPs standards and advise the proponent as to whether it is going to be considered an obstruction, a hazard, or have no impact on navigation. If it’s determined to be an issue, the FAA has advisory circulars that will inform the proponent of choices they could make (how far they could move it, lower it, mark it) if they wish. Ultimately, the FAA has no further say. Usually though a governmental body, city, county, state has to issue a permit and approve the project. And this is where hazards to navigation near an airport get stopped. Those local government folks do listen to what the FAA has determined and fortunately will usually only issue a permit for conditions that the FAA can through TERPS procedures address as, although not preferable, at least adequate. Two other entities also will frequently stop a proposed project if the FAA has issued objections. That would be insurance companies and the FCC if it is a radio tower. The FCC has to issue a license for radio type transmitters. If the FAA had determined that the tower will be an unacceptable hazard, then the FCC won’t license it. An insurance company won’t insure a project that is deemed risky by the FAA.

    • Which came first? The approach or the potato processing plant? I’d guess it was the latter with the approach being designed giving full consideration for the local obstacles. The minimums for the approach and MDA suggest a less than optimal but workable approach.

  8. Very unfortunate accident for all involved. RIP.
    It is always a bit of a stretch to analyze what might have happened based on a particular pilot action such as “taking selfies, solo, PIC, etc.”
    For now, I like to stick to more technical issues of the approach rather than trying to read the mind of Ms. Infanger.
    1. Age of the G1000. More recent versions of the G1000 (and in Garmin GTNs) this approach is a LNAV+V. While the “V” is advisory only and may not have been surveyed – it’s better than not having it.
    2. The approach min is 1 sm and the WX was reported at 1 sm. Even though the ceiling was broken 800 ft, with mist and snow, the visibility might have been less; especially if there was a “wall of steam” from the processing plant. Wondering if the ceiling was more like VV008. Of course, big difference between BKN008 and VV008. Both ceilings at 800 ft.
    3. Given the visibility, the runway or runway environment might not have been visible at 1 sm from the runway. But let’s assume that the runway was visible at 1 sm. Since the processing plant was about ½ mile from the runway, the airplane would have been at MDA or actually descending to the runway over the plant.
    4. Descent angle at MDA at 1 NM from the runway or from the step-down fix (JAMID) could also be a causal factor. At MDA of 4560 ft, and the runway elevation is 4152 ft – for a difference of 408 ft. Or let’s say from the step-down fix (JAMID) at 4800 ft to the runway (1.6 NM) the difference would be 648 ft. In either case, the decent angles would be 3.850 and 3.820, respectively. Not too steep but consider the rather short runway length of 4092-305 ft = 3787 ft including the displaced threshold of 305 ft and 80 ft wide. Not particularly short nor narrow for a Caravan but touchdown point needs to be close after the displaced threshold. Of course, what is key is not descent angle but rate of descent which is based on ground speed.
    5. I’m not familiar with the topography around the airport, but the missed approach is telling. A climbing turn is required immediately after starting the miss.
    6. It is always easy to suggest closing down an airport and building another one somewhere else – but what would be the cost? The easier action is to make this a circling approach with an MDH of 800 ft and 1.5 sm visibility. That may result in a few missed (delayed) FEDEX packages but better than fatal accidents or building another airport.

      • Sorry Kurt. LNAV+V will never show up on an approach chart. It is still an GPS approach flown to LNAV MDA. The +V would show up in some Garmin navigators and clearly stating that it is only advisory. Most recent Garmin navigators provide the +V (GTN and GTN Xis, maybe some G1000s). With GTN Xi and latest SW you will also see: LP+V, VOR+V and NDB+V

      • Advisory glidepaths – those annunciated with +V on the navigators – are not a part of the approach, and you will never see +V minimums published on approach plates. They are flown to the published LNAV MDA, but have vertical guidance.

    • Thanks for the complete and thoughtful reply.

      It makes me wonder though. Here we have a non precision approach with MDA approximately 400ft runway AGL presumably for visibility and safety. Yet on the approach path the aircraft is allowed to come within 200ft of the obstacle if at MDA.

      At my home airport there are two non precision approaches. Both have good MDAs. One 500 ft. or more above any obstacle on approach (LOC) the other 1000 ft. above terrain (DME ARC).

      Maybe KBYI just shouldn’t have an instrument approach if one cannot be designed with 500ft AGL MDA and at least that much clearance from obstacles in the approach path.

  9. A properly trained pilot would look at the approach chart and immediately recognize the reason for no published VDP, (obstacle penetrating the approach plane). An experienced pilot, (which 1000 hours is not) would look at that approach and the weather and immediately recognize the high risk associated with flying that approach to the published minimums, especially in a single pilot operation. There is a reason that minimums are called minimums – and there is no requirement to fly down to those minimums if conditions are not favorable.

    • Ron, the reason a VDP is not published is because it has not been surveyed, does not necessarily imply that there is an obstacle. Additionally VDP assumes a 3 degree glide path to the runway, in this case we know that the glide path is > 3 degrees.

      • A VDP is so often published on approaches with a 3 degree GP that we may by lulled (incorrectly) into assuming that a VDP implies a 3 degree GP. Not so. Take a look for example at the RNAV (GPS) RWY 8 at KGIC. A read of Section 2-6-5 of the TERPS informs us that there are multiple criteria that must be met to publish a VDP. The approach plate doesn’t tell us why no VDP was published, but the absence of a VDP certainly includes the probability of obstacles in the visual portion of the approach.

    • So shouldn’t the cargo operator have had specific Ops for this particular airport? ie. higher company-imposed minima until checkout? again, this airport and potato plant just didn’t appear – and I’m sure there have been previous low-vis days this operator flew into – this is very similar to Paul B.’s St Bart approach: a tricky approach the is best served by a checkout or special training – just because no one crashed here earlier doesn’t mean it wasn’t an accident waiting to happen – is the cargo company truly surprised here?

        • Similar in that once you get to MDA and have the runway in sight you have to do a very steep descent – it’s not a typical 3 degree, but rather almost 4 degrees. If you watch the Blanco report he goes into what it would take to go to MDA and fly until you see the runway environment and then try to make it in such a short distance –

  10. Well, since it seems to be my lot in life to tell ppl what they don’t want to hear, we have to consider another possibility: It could be that she wasn’t qualified. But because she was a woman, she was given the job anyway. (And, for the record, I look back at some of the things I did flying and confess that I wasn’t qualified.)

    Thirty years ago, before all the political pressure of today (quotas, affirmative action), we had a DE here who made 60 Minutes fame. (And so was infamous.) I had heard (as a CFI-I at the time) that he had passed a women after her IFR Checkride. The story was that she cried because she failed.

    We know how men are when it comes to women crying. (Myself included.) He told her that he would pass her but she was “never to fly IFR” until she obtained additional training.

    So let’s admit it – while it’s true that there are oft times barriers to women working in a man’s world, it is also true (and somewhat ironic) that oft times the standards for women are lowered. Already our Military, Police, Fire Depts have lowered physical standards to allow more women to “qualify.” It might have happened here too.

    • FFS. Men like you are the reason women have to overcome more to achieve the same status. Take your mysoginistic sexist caveman comments elsewhere – a story about an up and coming female aviator’s tragic accident is NOT the place for your spew. TBH, there’s not really a place, but this is the worst forum you could gave picked.

      Just in case you’re fuming at this woman who has the gall to call you out – I’ve been certified in 7 ATC facilities, managed 3 of them, and worked at FAA HQ. I have an ATP and a type rating, and I’ve been been in aviation professionally since 1991. I’m fairly sure I’ve got more aviation experience than most men and I’ve had to work harder than most men to get it. I’ve never had, or seen, or heard of, any woman’s standard being lowered for any reason – in fact, quite the opposite. You, my knuckledragging friend, are the reason that the standards are NOT lowered. Your attitude against women illustrates that counterpoint beautifully.

      Sit down.

      • thank you Allie for calling this person out – he and others just take pockshots sometimes that are just downright demeaning. They don’t realize the damage they do – the stereotype they help keep alive – it’s like they don’t have a mom, or sister or aunt or daughter that is affected by stupid comments – what it really comes down to is sour grapes cause they feel their dream has eroded and taken away by someone else, as opposed to just bad luck and/or lack of performance on their part – AvWeb needs to invest in AI to moderate these comments and the ones that blame everything on one political party or another –

      • I mean Era P – you’re really bringing this quota crap here – enough of your BS – a black FO jacks up an Amazon plane a woman crashes a Caravan and you come up with the same quota crap – I feel sorry for you

      • Mishap investigations need to take into consideration all possible contributing factors. So when Era P. offered the possibility that the mishap pilot was not qualified, but hired because she was female, are Allie M. and R N. really asserting that that situation never happens? Is there really NO POSSIBILITY that the mishap pilot was put in a position beyond her capabilities?

        Allie M., given your impressive experience, and simply being alive in our society today, I’m surprised that you somehow missed seeing women and minorities given opportunities based simply on their physical traits and not on their qualifications. As a former military flight instructor and current physician, to me, it seems to be the rule, rather than the exception. I’m also surprised that other than offering well-reasoned counter-arguments, you and R N. stooped to name-calling here. Simply speaking the truth, or considering a possibility, does not make Era P. misogynistic or “knuckledragging.” You may disagree with him, but keep in mind that applying repugnant labels to those with whom you disagree is a political tactic quite popular these days, and reveals your position to be quite weak.

        Consider this, Allie M. When I saw poorly-performing female flight students being given preferential treatment, whom do you think was most angry? The well-performing female flight students! Why? The bad ones hurt the reputations of the good ones.

  11. Single pilot IFR, difficult and workload intensive even in a Cessna 172. And last minute quick decisions on a short final in marginal conditions leave no margin for error. And, even very experience pilots do the last minute on approach screw ups also.

    While at BNA, mid ’90s, AAL B727 on a Cat II to rwy 2L, got slight visual indications on very short final. Capt. later said he did what he knew he had been trained for years to not do, he pushed over and dove for the lights he saw. In doing so, he took out the entire line of approach lights (fortunately frangible mounts). He didn’t even know he had done so. Because of vis, tower couldn’t see the runway. 30 minutes later, two departing aircraft reported the big mess of broken glass on the runway. FSDO investigated, found the damaged aircraft and tracked down the Capt. He and the crew swore they only thought they had done a hard landing.

    While tower at Clark AB, PI, mid ’60s, MATS C135 conducting a PAR approach in a TRW, on short final saw what they thought was the end of the runway. Was actually a perimeter road and a taxi driving on it. Even though the PAR controller told him several times, “too low for safe approach, go around, go around”, the Capt. dove for it, hit the taxi and short of the runway exploding. Load of troops and dependents.

    And just several years ago, FedEx Airbus arriving BHM early morning, bad weather, and a non precision approach, dove for visual and hit short of the runway killing themselves. The FO was a local young lady just recently hired. Had worked very hard doing all types of flying building time for this super exciting new job she had. The Capt was a very experienced guy. He made the decision.

    It can happen any of us. Just a one time brief distraction or bad decision. It has happened to us all in some form, I bet!

    • Unfortunately, barring discovery of some mechanical issue, the final NTSB report will itself almost certainly be based on a reasonable guess.

      Such as: Did she have the necessary visual clues to legally descend below the MDA and then lose them? That certainly happens, and no matter how good a pilot you are, it takes a measurable amount of time, time that is then often in extremely short supply, to make that transition from controlled eyes-outside VFR descent to controlled eyes-inside climb on instruments.

  12. I watched the Juan Brown (blancolirio) You Tube video analysis, and the accident was caused by descending below MDA in IMC. It will take the NTSB 3-years to reach that conclusion. The MDA is there for a reason, it you want to roll the dice and cheat it in IMC the result may be something other than a successful landing. The NTSB will never go into how frequent or widespread MDA cheating is in the freight operation business. I am sure the FAA will catch up with the practice some day when they start cross checking altitude reporting for aircraft on IFR flight plans with the ADS-B position on an approach.

  13. John, Is it not true that to land an airplane, at some point, it must go below the MDA? Otherwise we would never land during IMC!

    If it is true that the plant (chimney stack) is 1/2 mile from the runway – the airplane would have been past the step down fix and in an area where further descent would have been needed to land.

    Perhaps you might have intended to say: the airplane prematurely descended from MDA with the pilot not having one of the 10 items listed in FAR 91.175 clearly visible. In this case, probably the runway threshold or the threshold markings or the REIL as the other 7 might not have been relevant in this instance.

  14. The obstacle she hit was well below MDA and before the MAP.

    The only way she could have hit it would have been to have been 200+ feet below MDA before MAP in IMC.

    Had she flown the approach as published this could not have happened.

    These points are inarguable. You can’t hit a target you are safely above.

    Maybe there was a mechanical or other extenuating circumstance, but watching the well presented data on blancolirio it seems, at this point that this was pilot error.

    Nothing personal against her her family friends company or her sex (which is unrelated) but she was well below the step down and hit something.

    The take home lesson for all of us is to fear and respect MDA and MAP and GS.

    • But the pilot could descend below the MDA with the runway in sight. I seem to be repeating myself but one could easily have the runway in sight but not see an unlit tower in the runway centerline in a snow storm.

      Photographs that runway are pretty frightening.

  15. But wait! GPS approach is the future, let’s decommission all those pesky and costly ILSs…so says the FAA…they will rue the day they do that; US commercial aviation safety record was built on the ILS approach. (For any haters, yes I realize a 3.7 degree final descent path was probably never going to have an ILS, but making a point here that a lot of Gov’t money is wasted on things far less worthy than ILS systems.)

  16. After flying night freight for 39 years there is one thing not mentioned, fatigue. Fortunately I flew 2 and 3 pilot airplanes but more than once I’ve seen a crew member doze off on an approach or final.

    • I agree with Frank G. Fatigue is always a possible issue in any incident. However, in this flight, she took off from SLC at 7:00 AM and I believe it is about an hour flight to BIY in a Caravan – resulting in the accident around 8:00 AM. Of course, we don’t know when her duty time started and what was her rest profile 48+ hrs before the accident. I’m sure that the NTSB will be looking at fatigue also.

  17. Talk of the tower on approach reminds me of the concrete factory tower along the rail road on the approach into Peachtree DeKalb airport KPDK. The approaches had to be offset a bit to the left.
    There was icing, and I think she iced up the plane trying to get in there. I agree with the other commenter, I think she had the ‘I gotta get there’ pressure. And dropped in a little low with ice.

  18. As a former military pilot and instructor I understand that we all lose something when those we taught suffer tragic accidents. IAP’s are there for a reason and when we fly approaches within the constraints of the published minimums we eliminate these types of tragic outcomes. When you make a mistake or decide to push the minimums in aviation then you are leaving your fate up to luck. I think it is a mistake and careless argument to fault the construction of a facility that had the approach been flown adhering to the published minimums would have, at worst, led to a missed approach and diversion to a suitable alternate. May we all learn from the tragedies that will always be part of aviation.

  19. Meteorologist here. According to the weather report in the story, the temp was -3 and dewpoint -5. Author said Fahrenheit, but am assuming it was Celsius. Still, about 28 degrees and 1 mile light snow. So what were the runway conds/braking action? Other commenter said fairly short runway beyond displaced threshold. How low did she get on the first missed approach? Also, agree with speculation about possible icing, she may have spent a good amount of time in the soup doing the go-around.

  20. Am a CFII, lots of PIC time(not that much stick time since mostly instruction given). Anyway,
    looking at the approach chart plan view, we see that the step down to the MDA is just PRIOR to the obstacle. With the runway not being particularly lengthy, seems possible she made an aggressive descent towards the MDA, maybe even just a tad prior to the step down fix, had the runway/reil in sight and perhaps disregarded the MDA at that point. I also remind my instrument wanabees that an advisory glide slope(LNAV+V provided by the G1000) does NOT account for obstacles below the MDA. I advise them to ignore the “+V”. All this in light of the fact that altimeters read a touch higher in cold weather since altimeter settings do not compensate for temps, and, altimeters are “legal” as long as they are within 75 ft accuracy. Put that all together and we can understand that non precision approaches are the most dangerous during the “visual” portion.