When it comes to someone's personal attitude to risk, there is no arguing with that. Each to their own, and it cannot be objectively faulted.

What one can always debate is relative risks and how to minimise them.

The magic number which one could descend to on a DIY approach is a good one. 1000ft (the traditional UK MSA) is reasonable as one would expect the obstacle rules used in the production of the CAA VFR charts to protect you. How low can one go while still having this protection? I can't remember the exact rules the CAA uses to depict each bit of the terrain, and also if I recall correctly there can be uncharted obstacles up to 299ft above the ground.

Perhaps somebody can illuminate this.

The proper way to do this is to get out the 1:25k or 1:50k Ordnance Survey map and look at the contours very carefully. It should have man-made obstacles on it, and should be as up to date as the CAA chart which can itself be over a year out of date. If I was planning an approach I would use both charts, GPS+VOR+DME, and descend to 700ft above known obstacles. This is reasonable given that many NDB/DME approaches have a DH of that much but with far less accuracy in azimuth. 400ft is lower than I would do.

As regards a radar service in IMC, again I respect anybody's view that they would not fly in IMC without it but this hugely conservative attitude is unsupported by mid-air collision data (how many mid-airs have occurred in IMC?) apart from making IFR OCAS impossible most of the time. The most dangerous thing has to be trying to get into Wellesbourne or Stapleford on a sunny Sunday preceeded by weeks of bad weather; people cut you up on the inside, below you, above you, a total free for all at times.

In fact a very real problem in this department is caused by all those pilots who fly around without a transponder; this renders a radar service mostly useless because the ATCO is forced to report all horizontally conflicting traffic even if you are at FL070 and the non-transponding traffic is probably somebody at 1000ft.

Have at least an IMC rating, preferrably IR,

I hope nobody is assuming otherwise; if they do it would explain some of what has been written!

The magic number which one could descend to on a DIY approach is a good one. 1000ft (the traditional UK MSA)

I think that it would be a good idea to clarify what this means ? I think you mean 1000 AGL (or above highest obstacle in vicinity?)

I think you mean 1000 AGL (or above highest obstacle in vicinity?)

Yes; the traditional UK MSA calculation.

If I recall right (gave away the books recently) even Trevor Thom says this is OK, and he must be right! :O

I'd still like to know the rules of terrain depiction on the CAA charts.

There are I guess not that many procedural approaches outside controlled airspace in the UK. There are a whole lot more in the States. Presumably we may end up with more, if GPS approaches are ultimately approved.

Separation implicitly relies on pilots doing what they are told. There is clearly no backup so far as vertical separation is concerned whilst in the hold. I suppose ATC does have some indication of where the aircraft is from DF so that if for example the aircraft reports beacon outbound and isnt an alert ATC might spot a problem.

I recall one of two close misses I have had; whilst in the hold without radar. briefly I became VMC and almost immedaitely spotted another aircraft far too close for comfort. It turned out they were working Dunsfold (in the old days when Dunsfold provided a LARS) and yet had managed to be allowed to fly straight through the ATZ in IMC. Not surprisingly there was a rather terse telephone call between the two ATCs.

Sadly it seems to me the biggest risk is just this - other aircraft flying through the ATZ in IMC oblivious to the fact the "local airport" provides an apporach service. Also whilst the big sky theory is comforting unfortunately our arrangement of airspace provides far to many choke points.

Personally I feel 1,000 feet AGL is entirely reasonable, but no lower. From experience it is a hard IFR day when there are no cloud breaks at all to enable a visual descent albeit that does require a sound technique for maintaining a visual descent in a small break. I would feel that if you find yourself in such hard IFR conditions unless the forecast was seriously inaccurate you had no business setting off in the first place without having recourse to a procedural approach.

More or less as you say, just search for 299 here: http://www.caa.co.uk/docs/33/SRG_GAD_WEBSSL5.PDF

"Of the risks to be assessed in any DIY instrument let-down, I would have thought that the risk of collision with obstacles or terrain vastly outweigh the risk of mid-air collision."

Not sure about that, Bookworm! It is perfectly possible to plan an IMC approach taking into account the terrain and other obstructions. I am NOT advocating this to be done in the air - rather a proper plan in place with a survey done first! However, encountering another aircraft in IMC can be an unquantifiable risk.

Speaking as someone required to spend a fair amount of time IFR/IMC in Class G airspace, it is becoming increasingly worrying how many aircraft are IMC but apparently NOT talking to an relevant ATC unit for any kind of separation service. The really worrying ones are those not transponding mode C (possibly not equipped and therefore not supposed to be flown in cloud). They really are taking a big risk, if not in receipt of an ATC service (and therefore not communicating their altitude) especially those under the London TMA when it isn't possible to fly quadrantals. We seem to encounter them increasingly often; my most recent encounter was last week when the radar unit we were working advised of a "contact left to right, 2 miles, no altitude information". We already had the contact on TCAS but it had no altitude tag. We therefore reasoned that it was VMC beneath cloud. As it turned out, we were wrong! It was an R-44 helicopter flying in and out of cloud about 4 miles north of Denham. It passed only a couple of hundred feet below and almost directly beneath us, according to my co-pilot. As far as I know, there are no R-44s clear to fly IFR in UK...... not least because they are single engined!

The big sky theory regarding no mid-air collisions IMC generally works but only because outside regulated airspace there are often fewer aircraft in cloud than not! The small minority of pilots who fly in cloud whilst not speaking to a radar unit and don't have TCAS quite possibly believe there is a tiny risk simply from past experience, because they are likely to be blissfully unaware of other aircraft in their close vicinity - until one day when they will possibly be part of an accident.

The problem, ShyTorque, is that much of the time one cannot get an RIS (OCAS) and when you do the controller has to report loads of returns from non-transponding targets.

If Mode C transponders were mandatory that would be something else.

And without radar, ATC service is meaningless. "27 aircraft known in the area"....

If only as much time was spent by the training industry telling pilots to fly with Mode C ON as they spend getting everybody to pointlessly call up London Info....

HWD

That PDF you refer to is just a leaflet on VFR flight planning.

IO540,
The issue of VFR or IFR has nothing to do with what is charted and what isn't. Clarify what rules you are looking for.

I was looking for rules the CAA mapmakers use to determine the elevation steps at which they change colour, and what rules they use to decide whether to show say a 530ft obstacle and perhaps not a 500ft one a mile away.

Obviously they must have rules on which obstacles to show, when others already shown are in specific proximity. Looking at the chart, the relevant distance is probably a few miles - unless they are exceptional man-made obstacles in which case (say, two tall towers close to each other, like those easy of LCY) they show them all.

Amen to that!

[QUOTE=IO540]The problem, ShyTorque, is that much of the time one cannot get an RIS (OCAS) and when you do the controller has to report loads of returns from non-transponding targets.
If Mode C transponders were mandatory that would be something else.
And without radar, ATC service is meaningless. "27 aircraft known in the area"....
If only as much time was spent by the training industry telling pilots to fly with Mode C ON as they spend getting everybody to pointlessly call up London Info....QUOTE]
I agree with most of that, the use of London Info is often not the most intelligent use of R/T. I think some pilots use it as a "coverall" frequency to avoid having to speak directly to other airfields near to their route.
Unfortunately, the amount of LARS radar coverage in UK does seem to be decreasing with the demise of some military bases but .
The altitude details of a non transponding aircraft, where given to a more appropriate ATC unit can be passed to other pilots using the service. When this is done for me, I will change altitude if appropriate to give some vertical separation. Our aircraft carries TCAS but as you say, it's not a lot of use if the other aircraft don't have Mode C or don't use it (for the info of those who don't know, TCAS isn't so good in azimuth and can't be relied on to give positive lateral separation from a non-mode C contact).
The thought of meeting up with unannounced aircraft in cloud scares the pants off anyone wth a modicum of common sense. In the last five years this has happened to me on at least three occasions in class G, one of them under the London CTA. Two of these aircraft were known to the ATC and working the same unit as us. Both, having announced their altitude and "VMC" suddenly climbed into cloud without saying so, causing a conflict.... Both of these were "professional pilots" flying twins; one of them on an instructional flight and one supposedly watching the traffic. :rolleyes: I no longer subscribe to the "big sky" theory (neither does the UK helicopter pilot who has miraculously survived TWO mid air collisions).

I think you can get overly paranoid over TCAS reports.

The statistical picture for IMC mid-airs over the last 50 years clearly supports the big sky theory. You might get a dire looking warning of traffic say 1000m away but if your wingspan is say 10m (PA28 type of plane) and the airframe is 2m high then your frontal profile is 20 sq. m. The other plane is say also 20 sq. m. If it is 1000m away and at exactly the right level the probability of a hit is just the ratio of the wingspan to the horizontal separation, which is 1/100, but nobody can fly the height that accurately - even below the LTMA :) - so let's take a 200ft level difference (say 50m to keep the numbers easy) then the 1/100 chance improves by the ratio of 2m to 50m i.e. 25x i.e. to 1/2500.

So your dire looking TCAS warning was actually a 1/2500 chance of a hit.

So if the two planes flew those reciprocal routes every day, there might be a hit after 2500 days.

These are just back of a fag packet figures but they indicate the magnitues.

If this was on a common commercial-traffic route, such as one finds in CAS (airways) then you would have plenty of hits (esp. with the much more accurate autopilots in use, and the virtually universal AP usage during the whole en route and approach segment) but that traffic flies under radar separation. Hence the old joke about the job of ATC being to pack everybody into a small space and then keep them from colliding :)

I won't claim that the existence of CAS protects anybody much because a lot of people infringe anyway and if they are not Mode C (which they aren't either because of the expected strong correlation between "bimblers" and non-equipped planes, or because they turned the transponder off in case they bust airspace) then ATC has to assume they are OCAS below it. However, CAS must reduce the chances because most pilots do keep out of it - despite their wonderful map+stopwatch PPL training :)

When we get to OCAS, the traffic density is far lower. The UK is actually quite busy; you can fly the entire length of say Italy and not see or hear another plane. On a typical UK flight, say 200nm, one might see 5-10 planes but they are very far away.

The chances of a hit en-route in VMC is miniscule (a hit, as I show above, is not the same as a near miss, the difference in probability is of the order of 2-3 orders of magnitude) and much lower still in IMC.

In 500-600+ hrs of flying on my own I have been within 100m of somebody maybe 3x, both being below 2000ft. Assuming 100m x 50m separation is 1/250 that puts the risk of a mid-air for me at about 600x250= 1 in 150,000 hours and that assumes I fly in those places every time I fly, which I don't, not by a very long way. One was on right base to Le Touquet (with a French pilot having been cleared for a concurrent left base in French), one was N of the Isle of Wight at 1900ft, one was at 2400ft over MID. Only the last one might have been likely in IMC. I am not counting various "proximity cases" trying to get into Stapleford or Wellesbourne because that was VMC and if you go there you take your chances! Of course those are just the ones I know about but even adding a factor of 10 to it still gives a hit every 15,000hrs and that's with the same assumption (flying in the dodgy places all the time).

The figures indicate that one could fly around 24hrs/day and would still die of old age before hitting something. Just don't try to go to Wellesbourne or Stapleford on a nice day unless you can hack it with the locals :O

Of course I am happier with a RIS, VMC or IMC, but all those ********** without transponders make the service practically worthless much of the time.

Lies, damn lies and statistics :)

Seriously though, although the chance of a mid air is incredibly low, chances are that the two aeroplanes are exactly in the right spot. It is called Sods' Law ;)

I think there is a difference between shooting a home made approach somewhere you may not have been for a while, and one at your home field to get you back in after a days flying. Assuming you depart your home field in the morning in VMC, and there is nothing obvious in the vicinity (new chimneys and the like), then you're probably pretty safe coming down to 500' on your way back in. If you're luckily enough to fly from somewhere like shoreham, there is no danger making up a GPS approach which takes you out over the sea and descends you to 500' before coming inbound again, and I would argue that it is probably safer than the NDB approach for a number of reasons.

I wouldn't shoot a home made approach without first surveying it in VMC, though have carried out a number of home made SID's using GPS.....

chances are that the two aeroplanes are exactly in the right spot

Having just cycled for an hour along a very busy dual carriageway, I wouldn't like to guess which is going to get me first ;)

Life is not without risk.

I agree. The risks are fundamentally different in nature. What I was thinking was that if we announced a national "fly your own IAP" day to the GA community sometime, we'd probably end up with more aircraft in the sides of hills than we would aircraft entangled with each other. :uhoh: But flying a properly designed IAP is a low risk activity.

IO540, You seem to quantify collision risk differently to myself; each to his own.

My own reasoning is based on experience gained from flying in a professional capacity since 1977 and a few years privately before that. In that time, I have experienced a number of other "mid-air conflict" situations where I would probably not be here taking part in this discussion, had I not taken avoiding action, as in the two previous cases I mentioned.

One of those more recent "encounters" in cloud allowed us to view the expression on the other pilot's face as we banked hard away from him. He was head on at exactly the same altitude (same QNH) and he was alone in the cockpit and looked wide-eyed and terrified - I sincerely hope he learned his lesson. He passed S + L, very close indeed on our left side, presumably not having any chance to take any action himself. Inexplicably, he had just confirmed to ATC that he was remaining VMC and 900 feet below us.

The recent provision of TCAS (in the last eight years in my case) has proved to myself and my colleagues just how many times aircraft DO come dangerously close to each other, both VMC and IMC in Class G airspace.

Often, having spotted another aircraft in VMC, we notice the "other" pilot takes no avoiding action, when required to by rules of the air - which makes me think that the lookout/awareness of some pilots is inefficient. Unfortunately, a pilot who doesn't see conflicting aircraft, doesn't talk to ATC and doesn't have TCAS is likely to think that "other aircraft" aren't actually there and may be further inclined to be unduly complacent on another occasion!

I note that your stats take no account of the fact that aircraft fly between ground features in VMC and other waypoints when IMC. Whatever, the likely severity of the possible outcome surely means that merely relying on chance and/or statistics is unacceptable in aviation. Otherwise, having ATC and rules of the air are probably a complete waste of time....

BTW, I certainly don't see myself as paranoid, in fact I would be seen as highly irresponsible to take any other viewpoint. All I'm trying to do is safely reach retirement. ;)

I0540

What you also need to factor in to your calculations are the following

1 - Every man and his dog wants to fly at 2000 due to London TMA and high masts
2 - All tracking to and from same pinch points over beacons
3 - All routing around Hrow, Stanstead, Luton etc

I would suggest that your odds of a hit are now much improved:)

"And without radar, ATC service is meaningless. "27 aircraft known in the area"....

I dont entirely agree.

There is an increasingly held view that it is a waste of time talking to London info.

The fact of the matter remains there are significant parts of the country were there is no RIS or LARS and no other meaningful service.

Never the less I still find some of the traffic information that can be provided by London info or ATC useful.

For example, I can think of many occasions where the call is bla bla bla expecting the SFD at 10.45 at FL45. That is useful if I am going to be in that vicinity at a similar level at a similiar time. There are many other examples. If pilots give up talking to any of these service providers then all that information is lost and after all whilst flying in much of France is very peaceful it also seems pretty quite after a while :) .

That's another good reason to use a GPS in IMC. Even a handheld one (backed up by traditional instruments) will do the job of allowing you to route easily avoiding this type of hot spot.

Enroute IFR generally still has the protection of being controlled airspace. You operate procedurally, still with a system designed to stop collisions. Read my post again, I am not saying don't do it under any circumstances, just highlighting that there are risks involved.