Low Visibility Proc q 4 ATC & pilots

[EltorroLoco]

Greetings

Would be interested to hear the ATC & pilot thoughts on some questions regarding Low Visibility Procedures (LVPs) & RVR readings.

Recently @ OR Tambo Int we had LVPs in force which was correct wrt to the weather (using the eyeball), however the ceiliometer (excuse the spelling if it's not correct) indicated "No Cloud Detected" & the RVRs were all over the place. The transmissometers were indicating from 2000m to 200m and they change every 15sec or so.

What we did was, to not consider the met instruments accurate and still allow the aircraft to do a CAT1 ILS, since the ceiling was about 200ft (the tower is 200ft high) so they should get the field in sight & the terminal building was in sight from the ground which is about 1000m away. This is naturally not a recognised method of measuring cloud ceiling & RVR/Visibility. We figured there was reasonable assurance that most aircraft would land.

I have two questions. First, is this the right thing to do? Should we rather have closed the arrivals until the weather had improved sufficiently to be "genuinely" CAT1. I don't think it's dangerous to do a CAT1 ILS in CAT2 conditions since one just goes down to 200ft, see the lights = land, don't see them = go-around. As it were there were about 4 go-arounds out of 30 or so landings. Some of them were due to Tower giving the RVR readings, which were below some of the aircraft's company regs, some due to not getting the field in sight. There were no diversions however, all aircraft landed with only one having three attempts. My question basically is, would you be happier to attempt a precision approach in marginal conditions if the ATC thinks there is reasonable assurance of a landing or hold till they improve?

I'll be honest, Tower were not very happy with this plan, hence my post.

The second question is more for the pilots. Let's say your company regs state one needs e.g. 350m RVR to attempt an approach, and it has been more than that since you started the approach. The RVR readings update every 15s or so, so if the RVR readings go below this, do you want to know about it, especially if it says e.g. 325m?

Our SOPs state that the RVRs must be given once by Approach and once by Tower so one can imagine that the above scenario is entirely possible especially considering there are 3 readings on 03R and 4 on 03L, also which one is more important? My guess is obviosly touchdown (Alpha) and then roll-out (Bravo) with the others being less important.

Appreciation in advance...

[Spitoon]

My comments are from a UK perspective (and I feel that the UK is generally right when it comes to LVOs).

Unless there is some good reason to believe that the met data you were getting is inaccurate then you 'believe' what they tell you and act accordingly. If the RVR is u/s, pass the reported met vis to the aircraft.

What you can see with the eyeball is an accepted way of making weather observations (although in Europe it is becomming more common for automatic equipment to produce the data and both met observers and ATC sometimes seem to forget how to work without AUTO gear).

Don't know what the rules are in SA but in the UK we don't 'close down' arrivals in poor wx - we pass the met info and ask intentions, then try to accommodate the crew's wishes.

In the UK the rules are strict about RVR - this follows incidents in which the IRVR was though to be inaccurate (reading too low) because both ATC and pilots could see for miles but the slant visibility on approach was poor (and being accurately reported). This was in patchy shallow fog - don't know whether this is common in SA but it sure is in the UK in spring, winter and autumn!

In UK, rules are that RVR is given soon after contact with approach and then any changes are passed - this is often done by broadcasts and specific calls to aircraft about to start an approach. Sometimes a crew will ask to be notified only if RVR falls below a particular value - according to the book ATC should keep giving updates, but often don't to avoid adding to crew workload.