ILS Altimeter Checks

[Ejector]

I make reference to Jepp Terminal 2.7.3.1

Can anyone put more light on “unexplained discrepancy”, like how much error you would tolerate?

Please also confirm that
LOC & Glideslope is full scale allowed before missed approach
And VOR is half scale,
And NDB is 5’
And GPS NPA is half scale.

Thankyou

[Chimbu chuckles]

Pressure sensitive altimeters do not make allowance for ISA deviation and therefore Indicated Altitude = True Altitude ONLY on an ISA day. Since ISA days are as rare as rocking horse poop an altimeter will only read the published figure on the ILS chart (at the OM or DME check HT) that often as well....virtually never.

A pressure sensitive altimeter reads the height of a theoretical column of air at the top of which is the altimeter and the bottom of which is sea level....assuming QNH set not 1013.2mb...by measuring and comparing the local air pressure to the MSL datum pressure.

On a warmer than ISA day the column is taller (air less dense) and on a colder than ISA day it is shorter (air molecules more densely packed)

The magnitude of the difference between Indicated and True altitude depends on how far above the datum the altimeter is.

On warmer than ISA days the aircraft will be flying higher than indicated on the altimeter and as this is safer it has always been ignored in Australia because that is the predominant condition. This has led to a degree of ignorance on the whole subject in Australia. It follows that on a colder than ISA day the aircraft will be lower than indicated on the altimeter and this can be very dangerous but because the difference is very small at typical Cat 1 minimas there is a level of complacence.

Now consider the case of the ILS. The glideslope DME Ht check occurs at a predetermined distance along a, for all intents and purposes, angled 'surface' set a 3 odd degrees....say 5 DME/1500'. Lets 'freeze' the aircraft at that point. On a warmer than ISA day what will the altimeter read, higher or lower....and on a colder than ISA day?

By how much?

Can you reasonably do any meaningfull altimeter check without knowing?

Will the error be the same at the minima?

Lets look at the colder day as that is the dangerous one.

I have said above that on a colder than ISA day the TRUE altitude will be less than INDICATED altitude. Viewed from the pilot's perspective 'frozen' on the ILS above the altimeter is OVERREADING....it might be saying 1580' for instance on a really cold day. Once upon a time not so long ago the AIP prescribed adding that difference to the minima...you will shortly understand why that is dumb....the AIP doesn't say that anymore because at some point CASA learned how dumb it was....but I digress!

Clearly once you pass the OM or DME/HT check the only reference you have for height above the runway threshold elevation is the altimeter...at the 5 DME point above you WERE at 1500' (lets assume you were bang on glideslope and a MSL runway)
irrespective of what the altimeter said. As you approach the DH/DA your only reference is the altimeter...lets freeze the aircraft at the Cat 1 minima of 200'.

What is your TRUE altitude?

Here is a little 'rule of thumb' formula which will tell you. The most important thing to remember is the temperature datum is the airfield that gave you the temp via ATIS or whatever and that temperature as ISA devn is what counts NOT the ISA devn at cruise alt.

+ or - 4' per degree of ISA devn(ht/1000).

So applying this formula to the above ILS...lets assume temperature 0 degree giving an ISA devn of 15.

- 4 x 15(1.5)

-60 x 1.5= -90'

So if on the day of the above ILS it was 0 C on the ground and the altimeter said 1590' at 5 DME then the altimeter is accurate.

At the minima.

-4 x 15(.2)

-60 x .2 = -12'

At the minima if you descended to 200' Indicated altitude your TRUE altitude would be 188'.

If you followed the AIP as it was writ 20 yrs ago you would have descended to an indicated alt of 290' which would have been a TRUE alt of -60 x .29 (call it .30) which = 272' and possibly not got visual...you might even do it twice and then divert to your alternate for no good reason at all.:ugh:

You can see that if it was 30C on the ground the whole situation reverses and becomes safer...at the OM you would be indicating 1410' for a true alt of 1500' and at the minima indicating 200' gives a TRUE alt of 212'.

Now apply the same formula to a MSA of 7000' on a 0C day and how much terrain clearance do you really have? On a 30C day? And the MDA on an NDB or VOR or GPS approach?

Now you will be able to predict what the altimeter will say at the OM or DME/Ht Check before Topd and having allowed for known altimeter tolerance (checked against the airfield reference point at point of deparature) be confident that when you call the altimeter check on the ILS you actually understand what you just said and it wasn't just ignorant BS.

Hope this helps.

Chuck.

[Capt. On Heat]

Nice Chuck.

Also aside from the glideslope/dme check height points, general mental dme checks throughout the approach based on standard profile or advisory altitudes whilst not exceedingly accurate will monitor a safe approach. Remember NZ60 in Samoa after all. Better late than never!

[reynoldsno1]

The ILS check heights should also include compensation for beam refraction and earth curvature.... just to keep it simple....:confused:

[Captahab]

A common error is the wrong reynolds number set on the thronomister, this can induce a coanda effect.
Compliance with CAO 40.2.1, 3.2 and 3.5 tends to eliminate this problem especially if you glance at the dme at the outer marker.

[Chimbu chuckles]

Ignorance is bliss eh guys?

[Cloud Cutter]

I think the 'unexplained discrepency' is more in reference to 'false' or 'erroneous' glideslope (as in the NZ60 incident) as these are the errors that will kill you, and they will be far in excess of altimetry errors (which apparently can be explained).

Deviations from ISA pressure lapse rate will be of little consequence on an ILS (will only have an affect on DA, and the error at low altitude will be neglegable - you're only slightly above the point the QNH datum is taken from).

[Chimbu chuckles]

Cloud clutter...all true but surely one of the best ways to guard against false glideslope is to know fairly accurately what the altimeter should read at the OM? It is the last point on the ILS that you can positively check for all sorts of possible errors.

[Speeds high]

Checking the slope at the OM is a good safe guard against the false glide slope, however the only way to check the erroneous slope (as in W.Samoa) is to do at least 2 glide slope checks (1 could be the interception alititude followed by the OM).

The erroneous slope will show "on glide slope" at all altitudes on the ILS with no warning flags, so one DME check or Marker/altitiude check wont show this rare occurance.

:ok:

[Cloud Cutter]

Yeah true, but i've just never seen it cause that much of an error to be worth worrying about. The discrepancies you would experience on a false glideslope would be huge in comparison (unless you did the checks 1 mile apart which would be a bit foolish). To my mind the instrument error in many altimeters is far in excess of any ISA deviation error. I may be wrong, that's just how it seems to me. I always try to do the first check at glideslope intersept and the next at FAP or the published check hight. Most ILSs in NZ don't have markers associated with them, but they would obviousely be a good place for the checks. It gets tricky when using a non-associated DME which was nearly one of the holes in the cheese with NZ60.

[Weed]

GS/LLZ tolerance is half scale re 40.2.1 app 1 para 3.5 (c)

[Inka Dinka]

It may help some to think of the OM check as a 'vertical' positive fix.

The distance from the threshold (OM) or aid (in the case of a DME fix) is constant, as is the aircraft's height (by the glideslope). This allows an independent fix which confirms the servicability of the altimeter(s).

The 'unexplained' bit then is easily understood. If you are a bit high on glidepath and the altimeter reads, say, 30 feet high at the check, that's an explained error. If you are below glidepath and the altimeter reads 30' high, that's unexplained and may mean the altimeter is unserviceable, which would be nice to know as you got nearer the ground.


Inka

(BTW: do people agree that 'discontinue the approach' does not preclude one from continuing with a LLZ approach?)


PPS: Apologies Chimbu if you effectively said the above in your post. There were too many numbers and formulae for me; made me eyes glaze over :}

[Counter-rotation]

1) WRT "warmer than ISA is safer, colder than ISA is not" - I seem to recall a reference to -15 degrees in something I've read. Anyone else?

2) I thought tolerance for the ILS missed approach was full scale, from AIP ENR 1.5 ILS section (or Jepps I guess)? Does the CAO 40.2.1 reference above relate to the test requirements for I.R. issue?

Please note: I don't have my docs handy to check this, I am asking not telling. Also I am an ILS rookie....:)

CR.

ONLINE DOCUMENTS. Ah, they're a beautiful thing...;)

The reference I was looking for was (funnily enough)

AIP ENR 1.5; 7.3 Instrument Landing System - Altimeter Checks.

Why they talk about tolereance for LLZ and GP with reference to commencing the missed approach under THAT heading I'm not too sure. Note also it applies after the FAP also.

So I guess technically (WRT Ejector's original question) you are REQUIRED to commence missed approach (ILS) if:

1) after the FAP
2) LLZ or GP indicates FULL SCALE deflection

For other approaches, if, during the FINAL SEGMENT of the approach, tolerance is not maintained as required (AIP ENR 1.5 para 1.10 Missed Approach)
Sorry I don't know much about GPS NPA.

O.K... Where is this "tolerance" defined? The figures I've always been quoted (by those who supposedly know, and I wouldn't question) are:

NDB +/- 5 deg. VOR 1/2 scale DME +/- 2nm.

These numbers ARE written with reference to "when can I commence/continue inbound descent?". Also reference here to: ILS 1/2 scale:confused:

And of course, there's rules - and there's common sense.:ouch:

My question to those of you more skilled than yours truly -

Do you have a rule of thumb to calculate a revisedbDA/MDA for when you're flying a twin, heavy, that won't make 2.5% on one, if you hit the PUBLISHED DA/MDA and miss out? Surely you need to bug out before you get all the way down there? (Think PA31,, hot as hell, all seats occupied - I\'m sure you get my drift...):sad:

CR.