TCAS deep dive: DMOD, tau-cap and NAF-range

Imagine you are an air traffic controller in Madrid. You handle arriving aircraft during parallel approaches to runway 18L/R and on some days to runway 32L/R. Something is strange however: Since the approaches were changed from “dependent” to “independent” mode, you often get TCAS Resolution Advisories (RA´s) reported by pilots. Oddly enough, this only seems to happen when runways 18L/R are in use, but not on 32L/R…

…sounds too spooky? It turns out, this is EXACTLY what happened in Madrid. In this article we are going to find out WHY.

Figure 1: TCAS behaves differently during parallel operations.

The basic explanation does not “cut it”

TCAS is well-established in the industry and practically all pilots are familiar with its basic operating principle. We will therefore assume that the reader has a basic understanding of TCAS. In this context, the article refers to TCAS II, but we will simply call it TCAS for now. The aim of this article is to provide a better insight into the complex calculations that are performed by a TCAS computer. This will enable us to understand why on some occasions TCAS might generate RA´s during parallel approaches, even though it supposedly should not…

Time-based advisories, or not?

This is the concept that most pilots are familiar with: TCAS uses time-based “protection-zones” to generate the respective advisories. In this context, two parameters are often discussed: The Closest Point of Approach (CPA) and advisory threshold values, known as tau.

Figure 2: Time-based threat detection [1]

Using these time-thresholds, TCAS calculates protection zones around the aircraft and generates advisories accordingly.

Figure 3: TCAS advisory zones [1]

The advisory thresholds are then modified, based on aircraft´s altitude as indicated below. In order to achieve this, TCAS uses a parameter called Sensitivity Level (SL).

Figure 4: TCAS advisory thresholds vs. altitude [1]

The sensitivity is adjusted based on the height or pressure altitude respectively as shown above. This accounts for typical aircraft maneuvers and pilot reaction times. With this concept, the basic principle of TCAS can be understood.

There is just one problem: This is not quite how it works when two aircraft are flying close together.

What about parallel approach operations?

This becomes evident during parallel approach operations. Here we need some further knowledge to understand what is going on. For starters, it is important to know that ICAO recommends keeping TCAS in TA/RA mode even during parallel approach operations [1]. This is simply because the safety benefits of TCAS by far outweigh an occasional “nuisance RA”. There are new versions of TCAS under development, to improve its behaviour during parallel approach operations. These are known as ACAS Xo (see below). The interested reader is referred to THIS ARTICLE.

Figure 5: Future evolution of ACAS [1]. Refer to article linked above.

As evident from the example in Madrid, TCAS sometimes issues an RA during parallel approach operations, when controllers and pilots deem the separation sufficient, and the basic “time-based” logic certainly does not explain this. We shall therefore look a bit deeper into the TCAS specification.

Two tests: Range test and altitude test

During its normal working cycle (about 1Hz), TCAS runs a range test (lateral analysis) and if that is passed, an altitude test (vertical analysis). If both tests are passed, an advisory is generated. But what exactly does it mean to “pass” the range test?

If both aircraft are currently close or predicted to get close within the tau interval, the range test is passed. “Close” in this sense refers to a distance which is known as Distance Modifier (DMOD) [1] [2]. DMOD depends on the SL (see Figure 6) below.

Similarly, for the altitude test: If both aircraft are currently close or predicted to get close within the tau interval, the altitude test is passed. In the vertical domain, “close” refers to a vertical separation known as Zthr (see Figure 6). In the vertical domain, a reduced time threshold (TVTHR) is used, if the ownship is in level flight [1].

Figure 6: TCAS advisory thresholds [1]

The reason for the DMOD and Zthr is obviously that there is a proximity which is deemed unsafe, regardless of closure rate. The parameter ALIM is used to decide on the strength and the sense of the RA.

So far so good. But how come, in the Madrid encounters RA´s were generated even if the horizontal distance was greater than DMOD?

Enter: The Nuisance Alert Filter (NAF) and tau-cap mechanism

What to do, if two aircraft are converging and then turning parallel, so that tau increases and they are separated just a fraction above DMOD? It is obvious that any small maneuver could lead to a loss of separation. For this specific scenario, another “add-on” was developed for TCAS: The tau-cap mechanism. If the range test is passed and tau starts to increase again, its value is capped (not allowed to increase) until the aircraft are separated by a distance called NAF-range. It is a fixed value of 1.7 NM and has been derived from the highest DMOD value, accounting for typical aircraft maneuvers [2].

And herein lies the “sweet spot”: If tau-cap has been activated during the convergence and the aircraft end up being below NAF-range, the range test will always pass regardless of closure rate and even above DMOD. In this case, it is only the altitude test that prevents an RA from occurring. The instant the vertical separation drops below 600 ft, an RA is triggered (see Figure 6).

Crucial: Whether an RA is triggered during parallel approach operation does not only depend on what is happening “now”, but also what happened “in the past” (during convergence).

Pilots and controllers have no way of knowing if tau-cap is currently active or not. Obviously, this needs improvement and this is exactly what the industry is aiming to achieve with ACAS Xo.

Bringing it all together: The LEMD (Madrid) example

Figure 7: Situation in Madrid (LEMD) for runways 18L/R [3]

• Location 1

  The two aircraft track each other and the range test passes, as the miss-distance is predicted to fall below DMOD within the tau interval. There is no RA since the altitude test does not pass.

• Location 2

  The aircraft align with the respective approach courses but remain within NAF-range. Tau-cap is active. Still, the altitude test does not pass.

• Location 3

  As soon as the “high” aircraft begins descending, an RA is triggered, as the altitude test is now also passed.

  
  

The reason for the increased RA´s on runways 18L/R vs. 32L/R is the higher intermediate approach altitude on runway 18L/R which leads to TCAS using another SL. Furthermore, the runway spacing is slightly less for runways 18L/R than for 32L/R.

LEMD mitigation

After an immediate suspension of the independent parallel approaches following some 80 RA´s, the ANSP in Madrid has thoroughly studied the issue and developed slightly different approach segments, which should cure the problem.

Figure 8: Modified approach segments in Madrid [3] (they have since changed again)

Legacy routes are depicted in orange, modified routes in blue. This example goes to show how minor changes to arrival and approach routings can have unintended consequences. The swift reaction and investigation by the Spanish ANSP ENAIRE also demonstrates that solutions can be found even with today’s TCAS technology.

To finish this deep dive on TCAS, let us remind ourselves about a key point:

As pilots we shall always follow an RA and report it to ATC. The fact that we might “see” the intruder, does not relieve us from applying TCAS SOP.

References

[1]	Eurocontrol, «ACAS guide,» 2022.
[2]	Eurocontrol, «TCAS II RAs during parallel approaches due to tau–cap mechanism,» 2023.
[3]	ENAIRE, «SAFOPS #10: TCAS RA events during independent parallel approaches at LEMD,» 2024.