Simulating complex flying scenarios: Electrical failure in IMC

a flying simulation

With less than one mile of visibility, Lee on Solent, was invisible when Eugenio coasted in - Credit: Archant

For this second article in the simulator series we took inspiration from a recent ILAFFT (February 2021 issue). Eugenio Facci found himself in IMC with a NAVAID failure, while flying through icing conditions. The scenario here will be partially different: we will not have icing, but (distracted by an intermittent low volt warning) we will fly into deteriorating weather, eventually ending up in IMC without electrics. “Enjoy” the flight. 

BRIEF 
Aircraft: Cessna 172SP 
Flight: Shoreham EGKA - Newquay EGHQ  
Wind: 330/5 on the ground, 360/10 at 2,000ft 
Visibility: 6000 deteriorating 
Clouds: Broken at 2,300ft, becoming lower 
Scenario: Alternator failure, IMC, full electrical failure 

a flying simulation

This is what the same scenario would have looked like with 8km of visibility - Credit: Archant


It’s one of those days when an instrument rating wouldn’t hurt. I’m flying back home to Newquay after a work meeting in Brighton (in the fantasy world of sim flying I get to be whoever I want, and today I’m a successful businessman, proud owner of a Cessna 172SP). Such a flight along the south coast of the UK could be really wonderful, but today I am a bit concerned: there is a lot of haze, a cloudbase of 2,300ft (getting lower), and a northerly wind that is bringing deteriorating conditions towards my route. But for the moment I decide to cast my worries aside. In less than two hours I should be home, and that should be enough to avoid trouble. Also, there are a lot of airfields to divert to along the route, so I get ready to go.  

I take off from Shoreham Runway 02 but, on the climbout, I notice that the clouds seem to be lower than what the METAR said. In fact, I need to dodge a few of them while positioning for my initial leg. To avoid a number of Danger Areas and parts of Controlled Airspace, my route will initially track 250 towards Selsey Bill (which should be recognisable as a tip on the coast), and then straight west towards the north side of the Isle of White. But that is only the plan: I’m constantly dodging clouds, and it’s pretty clear that I won’t be able to cruise initially at my intended 2,100ft. So, when I’m about to pass 1,800ft I lower the nose, power back and level off for the moment, hoping to climb later. 

I do my top of climb checks and everything seems ok when, bling! The alternator warning light comes on. Hhhmm... Not great. I get my checklist for abnormals and start reading it but right away a full white cloud envelops the aircraft. I get ready to start my canonical 180 turn, but just as I’m ready to start banking, I exit the fluffy white. The weather ahead now looks ok again, though not great. And when I look again at the alternator light, it’s off. False alarm? 

The battery charge gauge seems to show a slight discharge (or is it on zero?), and that’s not really a help to assess the situation. I turn the landing light on, then off. There’s a slight movement in the dial, but nothing that brings conclusive evidence for my assessment. Maybe, I’m thinking, is there a problem with the light bulb itself? Or maybe some wiring is faulty? Suffering from home-it is, I decide to keep going. After all, there are many airfields in front of me where I could divert if the problem comes up again. However, to get a better view of my surroundings in the haze of the afternoon (is it getting worse?), I decide to descend to 1,400ft.  


Is the weather getting worse? 
For a few minutes nothing happens, although I am wondering if this is one of those days when the weather gets insidiously worse at a slow rate. But the pondering does not last long: I’m just a couple of minutes away from the coast tip at Selsey Bill when the alternator light comes on again. At this point Bognor Regis is the nearest airfield (behind me), with several others nearby – some with instrument procedures - so I decide to keep tracking 250 and follow dutifully the checklist. 

I get my ‘abnormals’ card out again, section ‘low voltage light’. The alternator circuit breaker is in, and cycling the alternator does not extinguish the warning. Not good. Next, I’m turning off the alternator switch, but the following item require some thinking: ‘Reduce electrical load’. I start with some of the suggested items: beacon, taxi, nav and strobes lights off. Then, while I ponder, a cloud envelops my Cessna again. This time, though, I’m more ready for a 180. I say my heading out loud ‘two-five-zero’, start the timer and turn left, towards the sea. A gentle one minute turn gets me back on track towards Shoreham, but it does not get me out of IMC.  

Most Read

The bad weather, evidently, is moving south faster than I thought. In fact, there’s a new ATIS at Shoreham that confirms that. Shall I turn north towards Goodwood? It’s nearer, and it does have a VOR that I could track. However, I’ve never been there, and from my chart it looks like they don’t have an ILS, while the ILS at Shoreham is one that I’ve flown a couple of times with my instructor. Can ATC maybe guide me down? More pondering... 

At this point, in the real world, I would call 121.5 declaring an emergency (“Non rated pilot in IMC”) and ask for vectors to Shoreham, along with vertical guidance (in this sim I can’t, and I will simply need to use my NAVAIDs). I then get back to my checklist but, in the heat of the moment, I read from the wrong section (‘excessive rate of charge’), and pull the alternator circuit breaker. A little more precious time is wasted in making sense of this action, and finding out the mistake. There’s a bit of learning to be gained here: I get one of my chart clips out and I use it to cover up the ‘wrong’ section of the checklist, leaving me to read only the correct one. I’m constantly looking in then out then in again, and it’s only a matter of time before I make the same mistake again. So, I’m ready to continue the correct checklist when, suddenly, everything turns off. The radio goes silent, the VOR is dead, and the GPS is blank. It’s a surprisingly sobering moment.  

An attempt to cycle the battery and the alternator turns out futile, while I realize that it’s roughly twenty-five minutes since the alternator light first illuminated. There was an alternator problem all along, I guess. Ok so, now, I need a plan to get back on the ground in one piece. First things first: I’m not instrument rated, so I make a silent commitment that all manoeuvring from now on will be smooth and gentle. Then, I’m no longer sure of my exact position, so I note down the time and turn south, towards the open sea. Shoreham, at this point, is just a few minutes away, and the temptation to turn towards it and ‘look for the airport’ is fairly strong. But it would be too easy to just end up lost and become another accident statistic.  

So, I decide to make a thorough plan, as always using my DODAR framework (Diagnose the issue, evaluate the Options, Decide what to do, Assign tasks, and Review the unfolding plan). The problem here is big, but pretty clear. I have no electrics (so, no NAVAIDs) and the whole south coast of the UK is in marginal weather or IMC. One option would be to continue flying south, descend over the sea until I’m clear of clouds, then head back north until I find the coast, and then fly towards an airfield. A second option would be to continue towards France hoping for a celebratory Champagne bottle upon landing, but my planning wasn’t as exhaustive as I would have liked and I don’t actually know what the weather is there. So, I decide to go with the first strategy: when I’m definitely over the sea I will descend until I get in sight of the water, down to a minimum of 300ft. I had the latest QNH when the electrics failed, so with that I’m sure that I will not descend too low. 

A quick look at the map, however, makes me aware of a ‘sustainability issue’ - pun intended. There is a large off-shore wind farm reasonably close to my estimated track, with blades reaching a whopping 400ft. They should be east of my intended track but still near, so to be extra-safe I decide to alter my heading from 180 to 225, and I note the new heading down on my nav log. There are no tasks to Assign here, since I’m alone and I can’t ask ATC for any help. And, finally, the first point where I will Review my plan will be at 300ft, if by then I am not in sight of the water. One last thing to add to my planning is fuel management. I had taken off at time 1447 with three hours of fuel. Now, with the electrics off, the fuel gauges are indicating zero, and I will need to estimate my fuel level. So, if all else fails, I will set a timer for thirty minutes of estimated fuel remaining (1717 zulu), and I will then do a forced landing wherever I am.

Time to execute the plan. I’m on a heading of 225, so I note down the time and start a slow descent at about 300fpm. It’s a bit of a thriller: one thousand feet pass without anything becoming visible, then eight hundred, then six hundred. I’m keenly aware that I now have less than one minute of ‘hoping’, when the greyish surface of the sea becomes suddenly visible at about 500ft. Phew... I level off, and get used to the idea that, for quite some time from now on, this will be my vantage point.  

a flying checklist

Eugenio inadvertently ended up reading the wrong section of the checklist - Credit: Archant

chart with clip on it

A chart clip, suitably positioned, blocked off the unwanted pages - Credit: Archant

More fatigue or more fuel? 
Levelling off makes one of the consequences of losing electrics annoyingly clear. The electric trim is gone, and in this sim there is no substitute manual trim (I remember the same situation in a real life Tecnam P-92). So, there is a trade off to be made between fuel consumption and fatigue. Ideally, I would like to fly slower (at best endurance speed), but I can’t trim for that speed, and so I would need to apply a constant pull back force on the yoke to fly at that speed. If, instead, I want to fly level without applying any force on the yoke, I need to fly at about 100kt and 2250rpm. At that speed I estimate that I have about two hours and ten minutes of fuel remaining now, which seems a decent amount. In addition, I remember Charlie Huke’s lesson about sitting comfortably, and the importance to do so particularly when flying IFR. So I decide to fly in trim at 100kt, even though that will cost me a bit more fuel. 

Being in sight of the water, it’s time to do a position fix, and then head north towards the promised land. Between doing a full DODAR and descending I have actually flown for a while on a heading of 225, and my position fix puts me thirty miles south of the Isle of White, roughly aligned in longitude (about three miles west) of the St Catherine’s Visual Reporting Point. So, at time 23, I turn right onto a track of 360. The wind at 2,000ft is exactly from 360 degrees, and so I apply no drift (I will only realize later that this is a mistake). In addition, while thinking about drift, I remember that I plotted my position earlier without taking the wind into account, and so I get my map out again to review my calculations. It turns out that the revision does not change the picture too much (possibly another mistake), and I should still see the land about two miles west of St Catherine’s VRP at time 40.  

Strangely for an emergency flight, the workload now becomes fairly low, and I have quite a few minutes to review what ‘the next event’ should look like (that is, reaching land). I notice that, coming from the south, I should encounter the south-west coast of the Isle of Wight at an angle of approximately fifty degrees, with the coast on roughly a 310-130 alignment. I am also aware that the encounter with the land will be sudden, and pretty quick. I estimate the visibility to be a bit less than one mile now, which means that at my speed I may have about twenty seconds between first seeing land and then flying over it. So, I decide that at time 35, a good five minutes before my estimate for reaching the land, I will put the map down and be super-vigilant to any sign of land. 

One thing that I do not notice, however, will become very evident a few minutes later. Within ten miles of my estimated arrival point over the Isle of Wight there are two peaks about 800ft high, with presumably land not sitting much lower in between. I’m flying at 500ft in low visibility, which means that I may not have much room to manoeuvre if I suddenly found those peaks in front of me. So, the lesson here is, to try and meet the land in a large area of low ground, and alter your course to do that is possible (in my flight that was possible, but I did not do consider it). 

So, at time 35, I put my map down and I start looking ahead, ready for any sign of land. It’s another piece of a fairly thrilling flight. I don’t actually know exactly what will be revealed to my eyes, and depending on what it is, I may need to react pretty quickly. At time 40, my estimate to meet the south-west coast of the island, nothing happens. Time 41, nothing again. A few seconds later, however, the magic encounter happens, and everything unfolds pretty quickly. But, wait, the coast is the wrong shape! It doesn’t run north-west to south-east, quite the opposite actually. And there’s high ground coming towards me! Where am I? Instinctively, I do a right 180 turn, and note down the time. I need to re-assess the scenario here. 

flying log written on paper

The improvised log that helped Eugenio keep track of his navigation - Credit: Archant

a map

His paper chart, marked with apparent and revised position - Credit: Archant

Plan over the sea, then check back 
Being over the sea gives me time to evaluate everything calmly. In fact, I realize, this is a good strategy to deal with this situation: do your planning over the sea, even writing down what you expect to see when you encounter land. Then fly towards land again, and see if it adds up. If it doesn’t, fly back out towards the sea and re-evaluate the scenario.  

In this case, during my ‘reconnaissance flyby’, I noticed that the coast is roughly aligned 070-250, and that there was ground next to the coast about 500ft high, or more. That looks like the south-east coast of the Isle of Wight, about five miles east of my previous estimate. Another possibility, instead, is that I’m very far from my estimated position, and in that case there are two more coasts with roughly a 070 alignment: the area around Bognor Regis and the area around Bournemouth, but neither should have high ground near the water. 

So, I’m reasonably certain that what I saw was the south-east coast of the Isle of Wight. To confirm that, I’ll fly back towards it, and then follow the coast turning left. If it is the Isle of Wight, I should be tracking roughly 250 for no more than four minutes, then 310 for about six minutes, then briefly west before a sharp turn right to track roughly 050. No other coast around here has this shape. Eventually, after leaving the north tip of the island, I will fly the same heading and, after a maximum of three minutes, I should be overhead the Lee on Solent airport, where I can land. If things do not add up, I can simply turn back towards the sea and re-assess the situation. 

So now, my plan well laid out, I’m ready for a second encounter with the land. I turn back on a heading of 360, estimating land at time 47. At the expected time land does appear, with high ground and aligned in the expected way, and I turn left to follow it. Tracking roughly 250, correct! After less than two minutes there’s roughly a seventy degree turn to the right, also correct. Long story short, the flight matches my expectations, I get a positive position fix, and at time 06 I leave the northern tip of the Isle of Wight on a track of 050.  

I cross the sea to the main land but, at time 09, the Lee on Solent airport is not in sight. Shortly after I’m over water again, and I follow my plan of doing a 180, get to the coast again and then follow it flying north-west. This time I do see the airport, its main runway approximately eighty degrees to my track. I learn something useful here: usually, when you do low visibility circuits, you know the airfield you are on, including the runway direction. In this case I do not, and on my first approach to the airport I try to eyeball the runway’s alignment, with the result that I end up forty degrees offset (easy to do, I guess, in low visibility). So on my second approach I line myself up with the runway, mark the direction on then CDI and say it out lout, and then finally fly a good circuit and land. 

map

And the sim’s record of where he actually was when he first sighted the IoW coast - Credit: Archant

Fragmented DODAR 
Unlike my first emergency sim scenario, where I learned most of my lessons during the flight, I find that quite a bit of learning is to be gained here after the landing, while reviewing what happened. The most difficult aspect of this flight, I found, was the quite fuzzy and constantly changing scenario that I encountered at the beginning (weather that appears to be deteriorating, while the low volt light goes off and on). This, inevitably, resulted in prolonged mental workload and in a fragmented DODAR (although I’d be glad to learn about better frameworks to deal with those kind of complex scenarios). It is the mental load, I believe, that contributed to several mistakes I made during the flight, including not spotting high ground near my route and an incorrect wind drift assessment.  

So, my first piece of learning relates to how to do emergency training: besides the traditional scenarios (PFLs, engine fire and so forth), I find it useful to train as well for these kinds of complex, unclear situations. They are a different ball game from the ‘conventional big emergencies’. Initially they are often much less threatening, but they do bring about an insidious additional problem: mental confusion. And it’s the latter that, in the end, may cause you to make deadly mistakes – the case of Air France 447 comes to mind. 

One potential, easy-to-apply rule that can help with complex scenarios is to systematically upgrade the severity of ‘yellow warnings’ to ‘red alerts’ when certain conditions are met – for instance, if you are not IR rated and the weather is marginal. The ‘low volt’ light is a yellow warning, not a red life-threatening situation. The checklist, in fact, does not immediately direct you to land at the nearest field, it asks you to work out the problem while in flight so that you can possibly continue towards your destination. In many cases where everything else is fine this makes perfect sense. But in a situation where there is something else potentially wrong (today it was the weather) such course of action puts you in a significantly riskier position. So, at the cost of following a false alarm, one possibility is to make a decision on the ground, before the flight, to automatically turn any yellow warning into a ‘land at the nearest airfield’ item when flying in marginal weather.  

A third helpful lesson is something that already emerged from our first sim emergency (Elevator failure, April issue): the value of patience and planning. More specifically, in low visibility without electrics, you will need to fix your position by looking at very few ground features, from very near and low, and for just a few seconds. It’s a very different type of visual navigation from the one I’m used to. So I found it useful to completely separate the ‘looking at the map’ phase from the ‘looking at the ground’ phase. Over the sea I looked at the map, setting out a precise description of what I expected to see when going back to land (eg “a six mile stretch of coast aligned 070-250, with high ground near the water”). Then I flew back towards the land, and confirmed (or rejected) my hypothesis regarding where I was, getting a position fix in the process. 

There is an additional action that would have probably helped me and that I did not consider (I guess because the mental workload rendered me a lot less creative). I could have lowered flaps to fly slower while still in trim, and that would have helped me in those moments when, flying over land, I was trying to recognize ground features to obtain a position fix.  

Finally, in a ‘IMC without electrics’ scenario the accuracy of your Direction Indicator (DI) becomes critical. There are two key aspects here. First, in some aeroplanes this instrument is electrically driven – in the one flown today it wasn’t and so it kept functioning, but you certainly do not want to follow an errant DI in this context. Second, in low visibility a one degree difference in your navigation may mean seeing, or missing altogether, your destination airport. So, as much as you can in a situation where you will be inevitably fidgety and stressed, try to do one thing at a time, and navigate precisely, with frequent compass checks, when it’s time to navigate.

simulated view from an aircraft cockpit

The view from the cockpit (left) when roughly one mile south of Lee on Solent, looking north - Credit: Archant

Simulator details 
The flight was flown using X-plane 10, a software that is almost identical to the one approved by the FAA for formal flight training. X-plane uses ‘blade element theory’, a process which allows for a realistic simulation of scenarios and effects at the edge of the flight envelope, including turbulence, stalls, the P-factor, and so forth. Hardware used for this simulation included control column, pedals, and throttle.

My top 5 lessons learned 
1) In IMC without electrics fly toward the sea and, when you are definitely over water, descend until clear of clouds 

2) Keep an accurate navigation log for position awareness 

3) When flying VFR towards marginal weather upgrade all yellow warnings to a ‘land at the nearest airfield’ action 

4) If you are following a checklist while constantly looking out and in, cover up the parts of the checklist that you do not need 

5) Be aware of the tendency to mis-use any IR training you might have in order to take on additional risks (even before the flight)
 

a map

Eugenio’s track to the airfield from the Isle of Wight - Credit: Archant

Instructor’s debrief
Congratulations on surviving another nasty malfunction, Eugenio. Overall, I think that you coped well, but as always there are a number of debrief points. Firstly, it’s crucial to point out that, given the prevailing weather, this trip should not be attempted as a VFR flight without a current instrument rating and a solid IR back-up plan.

Dartmoor, for instance, rises to over 2,000ft amsl on your track. Thus, even if not obvious from the ground, when you failed to get above 1,800ft on climb-out, you should have realised that achieving the 3,100ft safety altitude required after Exeter was not possible, and returned to Shoreham.

Relying on instrument procedures as a safety blanket in order to press on with a flight is concerning. Practising an approach with the comfort of an instructor while wearing a hood is very different to doing it for real, unprepared and down to minimums. Route study would have proven that the flight was not only impossible but made decision making easier once you had made the mistake of trying.

Moving on to the actual electrical malfunction. You were initially uncertain as to whether or not the warning light was functional. It’s ages since I flew a 172, but I would expect any modern aircraft to have a test function for any warning lights. I understand that in the sim there is no test function, but in the real aircraft a test would have shown you that the bulb was working, and what it looks like when illuminated. Having diagnosed a generator issue – and in light of the failing weather – I would have immediately turned back for Shoreham.

You also needed to be much quicker in load-shedding, and to consider more than just lights. Pitot heaters, landing light, transponder and fuel pumps (if fitted) are usually the biggest consumers, but so too are some modern glass cockpit displays. Knowing the capability of your battery, i.e. 20Amp/hr, and the current that each item draws will allow you to manage the battery. Maybe not in this case, but consider turning off everything, including the battery. This may allow you to make a long transit to a landing site, turning the radio back on to facilitate arrival when you get there. You’ll probably get thirty minutes from the average battery, much less with lots of transmitting. Manage it best you can.

At the risk of making this an abnormally long debrief, it’s worth briefly mentioning a couple of other issues. I wasn’t there, so it’s difficult to fault your decision making in a stressful situation. However, having achieved VMC below, I would have descended further. OK, you were only at 500ft, but coming down lower usually greatly improves visibility, as opposed to flying right in the cloudbase. It also greatly reduces the chance of inadvertently going IMC again. In my Wessex SAR days I have been forced to let down over the sea as you did. Having made landfall, I absolutely would not have lost that visual reference again by turning back out to sea, even if it didn’t quite match what I was expecting.

In this flight you could see the horizon, but almost invariably when you are over the sea in poor weather you’re forced to fly on instruments, whereas on land it’s a much reduced workload with something to orientate you. Rather I would have turned right, so that you could easily see land on your side of the aircraft. Flying along the coast would soon have revealed if it was the Isle of Wight. If it was, good! Follow the coast to a left base for Bembridge (route study again), or if you’ve hit the mainland, going right will take you to better weather and Shoreham!

Finally, your fuel exhaustion plan: excellent to see you considered this, but make sure that the plan includes enough fuel to fly north long enough so that you are over land when you might eventually force-land the aircraft. If you ditch in the sea without anyone knowing you’re there your chances are zero, crashing into the ground is also a terrible option, but in my mind better, as you’re more likely to be found!
Overall though, well done in creating a successful outcome from a very challenging situation – Charlie Huke