Antonov An-12 crash on approach to Lviv
On the 4th of October in 2019, an Antonov An-12 crashed 1,117 metres from the runway threshold at Lviv Aerodrome in Ukraine, killing five of the eight crew on board.
The Antonov An-12 is a four-engine turboprop transport aircraft, the military version of the An-10 although a little bit smaller (and a lot smaller than the An-225, the longest-bodied, longest-winged and heaviest aircraft in operation). The An-12 was nicknamed the “Black Tulip” during the Soviet-Afghan war, a reference to its transporting casualties out of Afghanistan in zinc-lined coffins.
This particular aircraft, an An-12BK, registration UR-CAH, had its first flight in 1968 with the Soviet Air Force. It was able to carry 100 fully equipped paratroopers or 20,000 kg of cargo. In 1992, it was transferred to the Ukraine Air Force. It had three overhauls before it was transferred to civilian use in 1995, when it was purchased by Lviv Airlines. The aircraft began operations at Ukraine Air Alliance in 2012.
Ukraine Air Alliance is a cargo airline based in Kyiv and one of the first private air enterprises in Ukraine to register with the ICAO. Their fleet consisted of seven Antonov An-12s but they lost three. One was destroyed in 2013 in a fire from an APU uncontained failure, one crashed into mountainous terrain in 2014, and in 2019, this one was destroyed after impacting trees and terrain on final approach to SE Lviv Danylo Halytskyi International Airport.
Lviv Aerodrome is located in the southwestern outskirts of Lviv and operates around the clock. They have a concrete runway 13/31 which is 3,305 metres long (over 10,000 feet).
There were eight on-board the Antonov An-12.
The captain was the pilot in command. He trained at the Kharkiv Higher Military Aviation School of Pilots and served in the Armed Forces of Ukraine. He had 6,570 flight hours and all but 20 minutes of that were on the An-12. I have to admit to being curious about those 20 minutes.
His first officer had 14,670 flight hours with 9,620 on type.
In addition, the flight crew consisted of the flight navigator, who had 13,385 hours (6,280 on type), the flight engineer, with 11,950 hours on type, and the aircraft radio operator, with 9,350 hours flight time (2,150 on type). It was a very experienced crew and they had all had flown into Lviv before.
A replacement flight engineer and two technical staff were also on board: an avionics technician who had qualified in 2006 and an airframe and engine maintenance technician who had completed his aviation education at the age of 47 and had five hours of experience when he boarded the flight.
The aircraft underwent maintenance in Toronto to deal with an indication failure on the landing system glide path beacon parameters on the captain’s side. The aircraft avionics maintenance technician disassembled the units and washed them with alcohol before drying, re-assembling and tightening the plug connectors in the navigation-landing device. The maintenance was completed on the 2nd of October.
The flight crew and two technicians flew the An-12BK from Canada to France, arriving at Toulouse aerodrome at 06:15 UTC on the 3rd of October. They delivered 1,537 kilograms of cargo. The flight crew completed their post-flight procedures and departed for their hotel. The technicians remained with the aircraft to “perform technical works”.
The next leg had been planned to Birmingham in England but the operator decided to change the route and scheduled the flight to Istanbul, Turkey, instead, via Vigo, Spain. The aircraft was loaded with 6,000 litres of fuel. The flight engineer was replaced but the rest of the flight crew was the same.
At 18:20 UTC, the An-12 landed at Vigo without incident. The crew did not leave the airport and did not have any opportunity for a rest between the flights.
While the crew was preparing for the flight to Istanbul, the airline’s dispatch service emailed a new flight plan from Vigo to Lviv, Ukraine, accompanied by the aerodrome information and the weather conditions.
The captain was very familiar with Lviv Aerodrome, as he’d served there as the pilot-in-command of the An-12 military transport aircraft of the Armed Forces of Ukraine. Everyone else in the crew had at least some experience of flying into Lviv.
The weather didn’t look great. For the estimated time of arrival, 03:45 the morning of the 4th, Lviv was forecast to have light showers, fog, broken clouds with a 60 metre ceiling (200 feet) with the visibility dropping to below 500 metres as the result of heavy rain and fog.
According to TAF, a wind of 270°, 4 m/s was forecasted for the Lviv aerodrome, visibility 3000 m, mist, broken clouds 210 m high; time (TEMRO) from 18:00 03.10 to 06:00 04.10 visibility was forecasted 500 m, mist, fog, broken clouds with a height of 60 m, broken cumulonimbus clouds with a height of 540 m; gradually (BECMG) from 6:00 to 08:00 04.10: wind of variable directions 1 m/s, visibility 10 km or more, no special weather phenomena, broken clouds with a height of 450 meters; time between 8:00 and 15:00 04.10: light shower rain, broken cumulonimbus clouds 600 meters high.
The report states that if the estimated time of arrival coincides with a forecast of deteriorating conditions, the lowest value should be taken into account when making a decision.
The captain was rated to perform landings on the An-12 with the following minimums for a CAT I landing approach:
- decision height: 200 feet
- Runway Visual Range (RVR): 550 feet
- CAT I ICAO
The minimum for a director mode landing for the An-12 is:
- decision height: 60 metres (approx 200 feet)
- RVR: 800 metres
The report says that civil twilight was at 03:51, thus dusk was falling shortly after the expected arrival of the aircraft. But Ukraine’s timezone is UTC+2 with sunrise at 07:28 and sunset at 18:55, so I’m not sure what is meant by this. The high intensity lights at Lviv were turned off at 05:26:25 UTC, presumably as the sky lightened.
The flight would be landing in the dark and conditions were forecast such that they may not be able to get into Lviv. Regardless of the oddity of dusk, it seems clear that the captain did not take the risk of rain and fog blocking the airport very seriously.
The flight information sent by the dispatch service included details of Boryspil aerodrome as an alternate if it was not possible to fly into Lviv. The forecast for Boryspil was broken clouds at 450 metres and a visibility of 10km or more. For the period between midnight and 0700, which covered the estimated time of arrival, the forecast showed light rain and mist decreasing visibility to 1,600 metres with broken clouds at 90 metres.
There’s no record of how much fuel was on board but they took on an additional 13,502 litres of fuel.
They also took on cargo, which was estimated as 13,000 kilograms of automobile spare parts destined for Bursa, Turkey.
The flight was delayed by two hours and twenty minutes but finally departed Vigo at 22:20 as flight UKL4060.
But there’s something wrong with the departure data on the FDR.
The Antonov An-12 manual states that the aircraft should “unstick” at 240 km/h with a take-off distance of 1,230 metres, after which it will climb away at 2.3 to 3.3 metres per second.
The flaps should be fully retracted once the climb is established, with an altitude of at least 150 metres and a speed of 310 km/h.
But that evening at Vigo, the An-12 unstuck at a speed of 272.3 km/h after a take-off distance of 1,600 metres, climbing away at a rate of 2.76 metres per second. The flaps were gradually retracted to 5° and only fully retracted at an altitude of 6,600 metres (21,000 feet).
About 14 minutes into the flight, a crew member asked about the fuel. Someone, probably one of the flight engineers, responded that they had 14,050 kg of fuel on board.
Based on this, the investigators estimate that they started with 15,000-16,000 kg of fuel on board.
The flight continued uneventfully, with the aircraft reaching a cruising altitude 7,200 metres (FL240) just before midnight UTC. At 01:20, they climbed to 7,400 metres (FL250) and then continued at that altitude until the start of the descent.
The EGPWS (Extended Ground Proximity Warning System) sounded only once during the flight, at 02:01:43 as the aircraft overflew Austria at 25,000 feet.
The flight crew first contacted Lviv Area Control Centre at 03:17. At about the same time, Lviv aerodrome, operating under low visibility procedures, switched on the high intensity lighting system, including runway edge lights, runway centre line lights, taxiway lights, etc. The fog had reduced the visibility at the touchdown zone to 150 metres with a vertical visibility of 50 metres (just over 150 feet).
The flight crew listened to the ATIS (Automatic Terminal Information Service) which was repeating information “Romeo”.
“Lviv, ATIS Romeo for 03:20. The ILS approach at the aerodrome uses low visibility procedures. Runway in use is RW-31. Runway surface condition known at 19:53 – wet, clear. The measured friction coefficient is 0.55. Estimated surface friction assessed as good. Transition level – 110. Warning: large flocks of birds in the aerodrome area and on the landing final. There is no wind. Visibility – 150 meters; visibility range on the runway at the touchdown point – 550 meters, in the middle of the runway – 550 meters, at the end of the runway – 550 meters, fog. Vertical visibility – 50 meters. Temperature + 3ºС, dew point + 3ºС. Atmospheric pressure QNH – 1013 hectopascals, QFE – 974 hectopascals. Weather forecast for TREND landing: visibility sometimes is 400 meters, fog; vertical visibility – 60 meters. Attention: the frequency Lviv-taxiing does not work, while taxiing, get in touch with the LvivTower at a frequency of 128.0 MHz. Please acknowledge receipt of Romeo’s information.
The flight crew confirmed that they had listened to information Romeo and were given clearance to descend to FL 120 (12,000 feet). The aircraft began its descent.
At 03:30, the ACC controller began radar guidance. He instructed the crew to descend to 4,000 feet and then to 3,200 feet. A few minutes later, the controller confirmed that the aircraft was 27 km from the LIV beacon (and thus from the runway threshold). He instructed the crew to take a left turn to heading 340° and cleared the flight for an ILS landing approach to runway 31.
An ILS (instrument landing system) uses directional radio signals to offer horizontal (localiser) and vertical (glide slope) guidance to the aircraft allowing for a direct descend to the threshold of the runway without visual aids. Inside the cockpit, the pilots can follow the localiser to align the aircraft with the runway. Often, such as on this morning at Lviv, the controller will offer radar vectors.
Having set up the aircraft for intercepting the ILS approach, the controller instructed the flight crew to report the localiser beam capture.
Once the aircraft has “captured” the localiser (aligned with the runway), the flight crew enters the glide slope, ideally while flying straight and level, intercepting from underneath the slope. Once the aircraft has intercepted the glide slope, the flight crew can begin their final descent to follow the glide path to the runway.
The An-12 was 15.7 km from the runway and 1,170 metres above the height of the runway threshold, descending at a rate of -4 to -4.5 metres per second. The aircraft completed the turn and entered final, still descending.
The crew reported capturing the localizer.
The controller instructed the crew to continue the ILS approach to runway 31 and reported that RVR (horizontal) visibility had increased to 800 metres at the touchdown zone with a vertical visibility of 60 metres in the fog.
The crew acknowledged the information as well as the frequency change to ATC Lviv.
The documented entry into the glide slope was expected to be 590 metres but the aircraft was actually flying at 660 metres. The An-12 was still 230 feet above the glide path. The captain increased the vertical descent rate in order to intercept the glide path.
The flight engineer extended the landing gear and then began to deploy the flaps.
The flight crew made first contact with Lviv Tower, reporting an ILS approach to the runway. The tower controller responded with wind information and cleared them to land. The An-12 was the only aircraft that the controller was dealing with.
As the flight crew acknowledged the clearance, they were 7.58 km from the touchdown point and now eleven metres under the glideslope. The An-12 continued to descend at a rate of around 5 metres per second, travelling at a speed of 290 km/h.
The crew should calculate the approach so that by the time of crossing altitude 500 feet (150 metres) above aerodrome elevation the aircraft should be in the landing configuration, balanced, at a stabilized approach speed, and the Checklist should be fully completed, except for specific items such as landing lights, windshield wipers, etc.
They were still 3 kilometres from the touchdown point when the An-12 passed through 100 metres, just 330 feet above the ground, and still descending.
The flight engineer called out “15, landing gear extended, landing weight 53.” The 15 means 15°, that is, the first stage of extending the flaps is complete.
During the distance of 4.7 to 4.4 km from the runway, the flaps extended to 35°.
A few seconds after the flaps finished extending, the external engine thrust was increased and the vertical rate of descent increased. Extending the flaps can cause a slight pitch-up and a decrease in the gliding speed, which can be countered with a slight nose-down adjustment on the controls and an increase in the engine thrust. The thrust of the internal engines was increased but then decreased again a few seconds later.
The vertical speed increased to 8 m/s, around 26 feet per second.
The situation was now dangerous. The aircraft was already beneath the glide slope and continuing to descend rapidly. The captain appeared to be attempting to increase the indicated speed, which was 256 km/h and dropping, by increasing the vertical speed. The aircraft continued to lose altitude. A few seconds later, the vertical descent rate increased again. They were now flying 65 metres below the glide path
Every ten seconds, the flight engineer reported the altitude (based on the radio altimeter). The navigator reported the distance to the runway threshold. No one was reporting the position of the aircraft relative to the glidepath.
At 60 metres (200 feet), the radio altimeter alarm sounded in the cockpit to signify that the decision height had been reached. They were still almost 2,000 metres (1,980 feet) from the runway threshold.
The Flight Operations Manual for the An-12 states that the minimum decision height for a director mode landing is 60 metres with a runway visual range of 800 metres. At this point, the flight crew must be clear of the cloud and have the runway in sight or break off the approach.
None of the crew members reacted. They continued to descend at a rate of 6 metres per second, while the indicated speed dropped to 237 km/h. The An-12 was 68 metres below the glide path.
At 1.67 kilometres from the threshold of the runway, the internal engines briefly increased and then decreased again. They continued to descend at 4 metres per second at a height of 48 metres (40 metres below the glide path). They were flying low and slow, travelling at 244 km/h instead of the recommended 270 km/h.
Analysis of the aircraft landing approach parameters at a distance of 11,000 meters, obtained from the recorders data, shows that the crew increased the instrument speed three times by increasing the vertical speed. So, at a distance of 8000 meters from the displaced threshold of the runway, the crew, after a slight loss of the indicated speed, increased the vertical speed to 6 m/s (established as 4 m/s), at a distance of 4200 meters from the displaced threshold of the runway, the crew, after losing the indicated speed, increased the vertical speed up to 5.5 m/s, with its subsequent decrease to 3 m/s, at a distance of 2000 meters from the displaced threshold of the runway and Habsolute=48 m, after the loss of the indicated speed to 237 km/h, the crew made a third attempt to increase the indicated speed for by increasing the vertical speed.
Forty-eight metres above the ground is just 157 feet. I cannot believe that they had the runway in sight or the ground in sight; someone would have reacted. Instead, they kept descending.
At a height of 30 metres above the runway threshold, which was just 20 metres over the ground that they were overflying, they must have come clear of the clouds. The elevator pitched up to 75% of its maximum value; someone desperately attempted to pull the An-12 up and away from the ground. It was much too late: the momentum of the aircraft ensured that it continued to descend at 6 metres per second. At almost 1,500 metres from the runway threshold at a height of 5 to 7 metres, the An-12 flew into trees and smashed into the high ground.
There’s a lot here so I’ll be looking at the aftermath and the investigation in a future post. Feel free to discuss your thoughts and even predictions in the comments.
The FDR for the takeoff makes me think the aircraft was more heavily laden than was on the manifest.
This whole thing smells dodgy. Theorising well in advance of the data: glideslope indicator failure that nobody wanted to admit, because of commercial pressures to get the cargo delivered?
I agree with Roger – the whole situation sounds dodgy. I’m also wondering how tired the flight crew was – but there were, it seems to me, enough crew on that flight that SOMEONE should have noticed that things were amiss.
One thing I’m curious about which I don’t known the answer to: Is EGPWS aware of the airplane’s vertical position relative to the glideslope? If not, is there something else on the panel that could/would trigger an alert? The only IFR landing I have ever watched from the cockpit was in a small plane with analog gauges, so I don’t know what happens in an airliner, but it seems like there ought to be a way to sound an alarm when the plane is close to the runway threshold and still not stabilized on the glideslope. i’m curious how that works in an airliner.
Not an expert so a little confused. What are the “internal engines”?
Thanks, Chris
That’s not a pilot issue so much as a translation issue, I think. They refer to the mode of the internal engines vs the external engines where I think internal is meant to say inboard. But I couldn’t prove it and I couldn’t find any simple explanation of how the four engines are used, so I left the term that the accident report used. Here’s a direct quote:
Sylvia, understood, yes inboard and outboard engines sounds likely.
Thanks again, Chris
Why does this AN-12 have windows in the nose cone?
From the picture of the wreckage, it looks like the 3 technicians outside the cockpit were the survivors.
The crew had that trans-Atlantic flight the night before. If they slept well during the day, they should’ve been ok. Had the technicians been able to sleep during the flight? Does the AN-12 have crew beds?
Why would the EGPWS sound on an aircraft cruising at 7400m when the highest point in Europe is 4809m?
They had 4 crew in the cockpit; the plate states the glide slope is 5.2%; someone should have had the mental acuity to figure out that that means ~200m altitude at 4km distance, ~150m at 3km, ~100m at 2km, ~50m at 1km. And obviously the ground can have a slope.
Why did they go to Lviev when the distance to Istanbul is similar?
Were they doing the pre-flight calculations based on a different load?
Were they using more fuel on the climb, and running out on the descent?
Was one of the pitot tubes (or the static pressure sensor) partially obstructed? so that the airspeed would show a lower speed on take-off? and maybe the altimeter showed a wrong rate of descent at landing?
Mendel: “Why would the EGPWS sound on an aircraft cruising at 7400m when the highest point in Europe is 4809m?”
AIUI, GPWSs will sometimes sound when overflying another aircraft. It happens particularly in holds where you have aircraft flying the same pattern directly above each other but, I suppose, it could happen anywhere.
The nose cone windows are a common design element in some Soviet designs of that era (1950’s). Originally meant for the bombardier and/or forward gunner ala WW2 bombers.
The design element carried forward for a while even when there was no bombardier, as in cargo planes. I think it was just a tradition.
I don’t believe the AN-12 was a derivative of a bomber design, but a purpose built transport.
By the standards I learned in the US, that’s a dangerously steep glide slope — ISTR 3 degrees is the max and some are at 2.5 degrees. More dangerous is how they intercepted; I was taught as you describe, to expect ATC to vector me into the GS from ~underneath (flying level until hitting the middle of the slope) and thereafter to ALWAYS stay at or above the slope. I don’t know what procedures these pilots were taught, but it sounds like they weren’t paying nearly enough attention to instruments; maybe everyone was staring out the windows hoping to see ground before they reaching decision height, given that the airport was on the edge of usable for Cat I?
I’m wondering about effects of fatigue; they had a long overnight flight, not really enough time to sleep given how hard getting to sleep in the daytime is (and allow for unwinding time, check and run-up time before departure) then a daytime flight of a couple of hours before another several hours at night.
And, like you, I’m puzzled about the captain’s time; surely there would have been dozens if not hundreds of hours in smaller planes before transitioning to a big turbine-powered aircraft. Did some of the log go missing and if so why?
The usual ILS glide slope is 2.5⁰-3.5⁰, with most at 3.0⁰.
There are a handful of published approaches with glide slopes even steeper than that (I would love to read a blog post about them–how about provoking “fear of landing” with a “dive bomber” approach?), e.g. Aspen, Lugano, or Lake Placid. And there’s a (published!) Space Shuttle approach for Kennedy Space Center with 20⁰!
There is a difference between the descent gradient and descent angle.
Both ILS at UKLL do have the standard 3,00° (degrees) descent angle -> so they are not “dangerous” (or steep) at all.
Steep approaches (usually anything higher than 4°) are not dangerous, but require aircraft certification and crew training. London City is a good example and they have an 5,5° angle.
In my aircraft we fly the steep approach with “steep approach mode” set in the FMS / EGPWS and 6 boards of 10 of the speed brakes extended.(SPDBRK extension below 500ft is otherwise forbidden!) There is a demonstrated height loss in the AFM for go arounds from the SA and thus the minima have to take that into account. However, the steep approach itself is not difficult at all IMHO – a typical overreaction of NAAs. I used to fly these apporaches in KingAirs without training etc, THEN the CAA of the UK came around and said we needed to have special training. Complete bollocks in my mind.
Perhaps you could explain the difference between gradient and angle; it’s not obvious.
However, I’ll note that there’s a lot of difference between a small airline plane (or a larger one designed specifically to be able to drop into small airports, like the late A-318) and middleweight cargo carrier.
And I’m wondering what you mean by “NAA”; if you mean “not an aviator”, note that I mentioned instrument training in my comment. I’ve been away from piloting much longer than Rudy (and never piloted anything larger than 4 seats), but I’m not a looky-loo or a whuffo.
CHip: NAA means National Aviation Authorities. I have never piloted anything with more than 9 seats, have an ATP and fly currently Citations and KingAirs for a living. (which does not says ANYTHIMG about my knowledge or capabilities, btw)
I guess I don`t have to explain what an angle is ?
The gradient is sorta the same thing – its the horizontal distance traveled (Groundspeed) vs change of height. Which, if you´d draw it, would give you an angle.
A descent ANGLE of 3.0° means a gradient of 5.2%.
E.g. if you travel at 120kts GS, you´ll need a descent rate of 634ft/min to maintain an gradient of 5,2% – > that is an angle of 3.0°
So, what I mainly wanted to say is, that UKLL / Lviv is a STANDARD 3.0° approach on both ends. The number 5.2 led you to believe its an steep approach – which it is not. I have landed in Lviv about 40 times, just sayin….and I have the Jepp plates here.
About the capabilities of an AN-12 ? I´m by no means an expert, but I have ssen them fly and they are in essence a military transport the Hercules (C-130) or the Transall. They all can fly fairly steep approaches, I would venture a guess here and say that they might fly steeper than anything “civilian”. Those huge props in idle will brake an awful lot. A KingAir B200 with 4 blade props has a never exceed of 259KIAS – on a 3° ILS you can keep this till 4 miles and be at Vref before the numbers just by idling the engines and putting the props in fine pitch. And that is basically what you do in an emergency descent and there you can build descent rates of appr. 6000-8000ft / min. Add gear and flaps and you overtake any Bechstein Piano thrown out of a highrise…. the only “trick” is to known when to apply power in order to keep from smashing into the ground….
Great writeup, and great comments, I look forward to future installments. I’ve never been a pilot, but that such an experienced crew could do this, something sounds off.
I have been looking at this write-up and must confess that I find it difficult to make much sense of it.
I have been in Lviv quite a few times, with the Turbo Commander as well as the Stallion (converted and upgraded Citation).
The runway was very long and the glideslope was not 5.2 degrees, that would be if the final approach is at a very steep angle due to surrounding obstacles or terrain.
There are some confusing issues, like “internal and external engines”.
I forgot which type of aircraft, but there has been a type with 8 engines mounted in pairs, driving counter-rotating propellers. So that configuration looked as if it hade four, rather than eight engines. But the AN-12 is a four-engine aircraft. The clarification of “inboard” (nos 2 and 3) and “outboard” (nos 1 and 4) engines is the only explanation that makes sense.
The operation of the flight seems to have been a bit haphazard. Like a cargo ship, where the nest destination depends on the available cargo, this aircraft went all over the place without much of a plan, let alone a schedule.
I agree that the aircraft may have been well over its maximum weight. To what extend was it within limits at the end of the flight?
I heard that in the old Soviet Union the captain could have been the overall manager of the flight, not necessarily an active pilot. In my air cargo days a Dublin-based operator leased one or two small Antonov aircraft – I think they were AN26. Smaller than the Fokker F27, they nevertheless had a crew of 5. The commander did not fly the aircraft. Maybe this explains the odd hours logged by the captain of the flight that is the subject of this blog?
It seems that the crew were suffering from fatigue. Another contributing factor – I am speculating – is that a steep approach, though by itself not dangerous, does require more attention because, if the aircraft is low AND heavy, it has more downward momentum. If in addition it is overweight on take-off, the crew had incorrect figures to compute the reference speed. The aircraft may have been heavier than it should have been, flying at too low a speed due to excess weight and had a high downward momentum.
Combined with a crew that were possibly tired and you have the recipe for disaster.
But of course, we are speculating.
Mendel and Chip come
… Mendel and Chip come close to what I think happened.
Anyway, I had a look, or tried to, at the approach plate.
Unfortunately, the resolution does not allow me to enlarge it without losing definition, but I am nearly sure that it shows a 3 degree glide path.
The report says that the glide slope was intercepted from below. Which is as should be. The crew had plenty of time to get “in the groove”, to set up a stablised approach. But they did not, engine power was changed, the vertical speed was increased several times. The aircraft became dangerously low, below the glide slope. Possibly, at low speed and high weight. the aircraft’s downward momentum was high. When they finally realised how low they were, engine power was not enough to arrest the descend into terrain. Maybe the crew suffered from fatigue and did not realise the danger. During the approach, already low, the aircraft still descended further below the glide path. This is another strange fact, if I read correctly, the gear was extended, flaps moved to what seems to be a landing setting. Warnings sounded. It was all too late. Mendel speculates about an instrument failure, like a static port? Absolutely feasible. And what was the cargo? Was there something that released toxic fumes, maybe military grade stuff, leaking into the cockpit? But surely most of that would release a suspicious smell.
A strange case.
Reading a second time, I realize I didn´t explain very well….
A gradient of 5,2% means, one has to descent 5,2% of the distance travelled horizontally.
Eg: 120Kts GS are about 12152 ft in one minute. 5,2% of that are roughly 632ft per minute required descent rate.
If one draws a triangle, the adjacent line is the 12152ft, the opposite, sitting at a 90° angle is the 632ft and the hypothenuse sits at a 3° angle to the adjacent line.