In a tragic case that demonstrates that carburetor icing is not limited to cold-weather or high-altitude operations, the National Transportation Safety Board’s (NTSB’s) final report on the May 14, 2022 fatal Cessna 172 accident on a bridge in Miami concludes that carburetor ice was the probable cause. The pilot, who was also an air traffic controller and Airframe & Powerplant mechanic, died in the accident and his two passengers were seriously injured. At his most recent medical examination in 2020, the pilot recorded 149 hours of flight experience, some 101 of which were in the make and model of the accident aircraft. His two passengers, both family members, were seriously injured and five motorists suffered minor injuries.
The pilot initially took off from Miami Homestead General Aviation Airport (X51) just before noon and landed at North Perry Airport (KHWO) in Hollywood, Florida, about 12:20 local time to refuel and pick up the passengers for a sightseeing flight to Key West International Airport (KEYW). At 12:48, as the Skyhawk was flying along the coastline at 1,200 feet under flight following, the pilot issued a Mayday call to Miami Approach: “…lost engine power, uh looks like um…let’s see…I don’t know where I am going to put this down but I’m going down.”
According to the pilot’s family, one of the passengers could not swim and there were no flotation devices on board. They were not required for the route of flight. “This may have influenced the pilot’s decision to land on the bridge rather than the water,” according to the NTSB.
The pilot told his passengers he was going to land on the Herman B. Fultz Bridge, which spans the Haulover Inlet. According to the NTSB report, the bridge is about 1,257 feet long and 56 feet wide. “The edges of the deck were equipped with sidewalks, railings, and streetlights,” the report reads. “The bridge was configured with four traffic lanes (two northbound and two southbound) separated by a raised concrete median.”
The NTSB reported that the northbound landing Cessa struck one vehicle from behind and another from the oncoming southbound lane, before coming to rest and bursting into flames.
At the time of the accident, weather at HWO (nine miles northwest of the site of the crash) showed an air temperature of 29 degrees Celsius and a dew point of 19 degrees, yielding a relative humidity of 55%. Also, according to the NTSB, unofficial reporting stations near the accident site recorded relative humidity levels between 63% and 66%. The post-crash examination of the wreckage showed the carburetor heat control was in the “off” position.
The NTSB report stated: “Weather observations indicated that the relative humidity exceeded 50% in the area of the accident site and along the route of flight, and review of atmospheric soundings revealed conditions conducive to the development of moderate carburetor icing at the airplane’s altitude. Based on the available information, it is likely that the pilot’s failure to use carburetor heat resulted in an accumulation of ice within the carburetor, which subsequently resulted in a total loss of engine power.”
Honestly, the NTSB needs to do more work. It seems like carb ice is their recent go-to with way too many accidents. If 65% relative humidity is the fatal range, we would be seeing C172 dropping out of the sky like flies every day. Apparently no note was made of the throttle settings at cruise.
Agreed, a friend recently was involved in a crash (no fatalities or serious injuries) and this is indeed their default answer for engine failure when they can’t find any other cause. The thing is, post crash, there is no way to prove it was carb icing. So for an agency that is so structured to focus on demonstrable facts, pointing to carb icing with no positive proof of that as a cause is irresponsible. Worse, if the pilot survives to fly another day, a finding of pilot error like this makes it very hard to get insured apparently (my friend is finding this out right now).
“The thing is, post crash, there is no way to prove it was carb icing.”
If by “prove” you mean find direct evidence of an iced up carburetor, you are correct. But engine analyzers and inspection of the engine post-crash can strongly suggest carb icing as a likely cause (I don’t know if that’s the case in this particular accident, though).
How do they know it wasn’t water in the fuel lines vs ice?
Very true. I agree that it is very likely water ingestion could be the culprit. If sampling the fuel isn’t done with the aircraft perfectly level then the slight dihedral of the wings will be negated and a slug of water could easily avoid detection until the aircraft is airborne and the engine quits. Piper Cherokees are much less prone to this due to the pronounced dihedral and that much less water can collect near the underwing drains.
You correctly pointed out that there is no way to prove it was carb ice, then scold the NTSB for pointing to that as the probable cause. How would you handle the decision?
ADSB might give a clue as to the throttle setting via the sightseeing speed at which it was traveling.
It’s not just the relative humidity that matters, though that is a big part of it. Throttle setting, as you allude to, is also an important factor, as is the total amount of moisture in the air (90% humidity in the winter could be less total moisture content than 40% humidity in the summer).
I have had a few cases of carb icing (in Lycoming-powered PA-28s, which have a low susceptibility to carb icing), in one case almost severe enough to cause engine stoppage before I realized what the problem was and applied full carb heat. One can fly for years and never have a problem until one day you fly in just the right conditions for it to form.
Carb heat does work…as long as it’s used while the engine is still running.
Prior to reducing power for a descent and approach the carb heat needs to be selected full on. 10 to 15 seconds later reduce power and begin your descent. leave carb heat on until short final then don’t forget to tie down when parked. Most problems avoided!
That’s good for approach and landing, but that’s not the only time carb icing is possible. In fact, my worst carb icing took place at cruise power.
I’ve also seen EGT rise while carb ice was forming at cruise. Go figure.
The accident aircraft was an O-300 Continental powered 172H, which is more likely to encounter carburetor icing than the 1968 and later models with O-320 Lycoming engines.
Mark, I wouldn’t write “Unlikely” since this is both a well-known risk, documented in the POH, and trained into PPL students, plus it did actually happen. “Surprisingly Mundane” might better capture what the headline writer was reaching for.
These days, I find that pilots are lackadaisical about Carb Heat on carbureted engines.
Carb Ice is very unusual on Lycomings due to the carb being attached to the oil sump and the induction air passing through tubes in the sump.
Continental engines have a more efficient cold air induction system, and are more subject to Carb Ice.
With Lycomings or fuel injected engines being ubiquitous, perhaps pilots are not being properly trained to be aware of Carb Ice on carbureted engines.
An ATPL with extensive time in everything from single seats to heavy jets, I currently instruct on antique airplanes, and find a lot of current pilots relevantly ignorant about Carb Ice possibilities on carbureted engines.
“In a tragic case that demonstrates that carburetor icing is not limited to cold-weather or high-altitude operations”
That’s a rather bizzarre statement, since carb icing has always been MORE susceptible in warm, humid conditions than cold, dry conditions. Carb icing must not be confused with airframe icing.
An O-300 should give a pretty good advanced notice before actually quitting….that should especially alert an A&P. I’ve had numerous carb ice issues with Continentals. The first one the engine was so rough that you could feel it through the airframe. Myself and anther pilot sat there brain dead until the engine quit. The next time I quickly became a firm believer in the carb heat knob. I used to fly a T-34 with the original 225hp O-300. It had a carb heat gauge. I was always surprised at how often the carb temp was close to the freezing level. And yes, warm humid is a good time to need carb heat.
Exactly what I was just thinking … why not have a thermal sensor?
Years ago I was a student pilot and ballast in the right seat of a Cessna 150 for a friend that was working on getting his commercial ticket. It was a warm sunny day in the southwest. As we were on one leg of cross-country flight I noticed the instrument panal vibrating. My friend was fiddling with fuel mixture and throttle as we developed a mild sink rate. My thought was we had carb ice or we were going to land in trees no matter what, and he needed to sweat a little longer before I suggested carb heat so he would become a better pilot. He went on to become a Airline pilot and retired years ago. Cccasionally he would remind me he would have landed in the trees if I hadn’t been long.
The PA 28 Dakota with the 235 Hp engine is very prone to Carb Ice. I had it many times. A small carb on a big engine is a prescription for carb ices and there is a large pressure drop after the carburetor thus the gasoline can evaporate and cool the air enough to form ice. I got carb ice many times in regular cruise settings. in that aircraft operating it at less than full power means pull on carb heat under almost any circumstances.
It happens and people need to be VERY aware.
Use carb heat anytime there is visible moisture, and or the humidity is high and the throttle is anything other than full open. Carb ice is unpredictable, use the carb heat.
Experienced card ice one time flying in the rain, power drop was immediate, almost killing the engine. Applied carb heat immediately and engine roared back to life. Years later I turned back to the ramp with a student because of no RPM drop with application on carb heat during pre takeoff check. One of the other CFIs mentioned, “oh, you don’t really need that.”
I hate to be a Monday-morning quarterback, and I don’t know what the pilot was seeing, but it seems to me that attempting to land on a bridge is almost always a very bad idea. Looking at the picture, I don’t see how it could have been done successfully, even without traffic. But there was traffic, and a bad outcome seemed inevitable. I understand that both time and options are very limited at 1200 feet, and being a low-time pilot probably didn’t help. I’m glad I’ve never been faced with the same situation, though, as I can’t guarantee I would have done any better.
The RV-12 I built in 2013 only offered a carbureted Rotax. I called them and complained that a fuel injected Rotax existed and why couldn’t I buy one from them. They explained that it hadn’t been ISO approved and if I wanted a Rotax I was stuck with the 912 carbureted engine. Since I had no desire to have a fuel system that required choking, carb heat etc. I opted out of purchasing the Rotax from Vans. I bought a modern fuel injected 3.5 liter engine which ran perfectly for the ten years that I flew my RV-12 (high altitude mountain flying, high temp desert and low temps, high humidity never a problem. Because of this choice I did my own engine wiring and sensors. Learned a lot about engine installations and performance. This modern CNC machined engine came with dual ECU’s and FADEC and ran flawlessly, carburetor ice never a problem, clean proper fuel injection, altitude computer adjusted. clean plugs. The only problem I ever had was having to use 100LL at airports (another source of angst) when I could burn 91 octane lead free pump gas from my local Circle K at half the price. Cleaning the gray lead from the plugs and an oil change for using lead on those cross-country fights. I can’t remember the last time I bought a vehicle with a carburetor or ever asking my Toyota dealer if they offered a carbureted system for a Tacoma, never. GA has been in the Dark Ages for decade after decade. 100LL, Magnetos, updraft carburetors and the list……..but we’ve been through all this on many other posts, and yet we trudge along through the miles of FAA paperwork clinging to classic “Vintage” aircraft engines designed in the 1950’s.
My thoughts exactly. It should not be too difficult to retrofit electronic fuel injection — you get kits for old cars where carburettors mean they are banned from many areas because of exhaust problems.
You could even combine it with systems where the spark comes from a light box, and not a heavy, magnet laden wheel spinning around.
But mention that at the airport and people turn purple with rage…
If you fly an airplane with a carburetor, adding full carb heat should be your first reaction when you see a drop in RPM or manifold pressure. It should be instinctive, just like lowering the nose to recover from a stall.
I have experienced 2 instances of carb icing in several thousand hours – one in FL and another in CA both requiring a few minutes to analyze, confirm and then resolve with carb heat. More recently in a flight from FL to MN, I observed what I thought was carb heat but could not resolve so made precautionary landing.
The rubber like gasket material behind the Challenger intake air filter had deteriorated over several hundred hours and been ingested clogging the carburetor venturi. Makes me wonder how many others have also overlooked the annual ICA concerning gasket replacement?
I experienced carb ice once in a 152 while on a cross country. Engine started running very rough and I pulled the carb heat out and the engine almost quit running completely. I pushed it back in half way just so it would keep running and started looking for a place to land. In about 5 minutes it cleared and was back to full power. I think if I’d pulled the carb heat out all the way it would not have kept running. Not in the procedures but sometimes you have to pay attention to what’s happening and adjust accordingly.
If only there was a possible solution to eliminate carb ice. Ideally a solution that improvedd efficiency and power delivery that has already been tried and tested on the same engines for years would be ideal.
I’ve owned two C-182’s, the first a 1959 “B” model and the second a 1965 “H”. Unlike the later H model, the B was a great trainer for dealing with carb ice. The difference between the two was quite dramatic and was doubtless the reason the B had a carb air temp gauge, which I believe might have been standard equipment.