Exit Separation Revisited

Exit separation has become a point of contention at many DZ's lately. Years ago, when belly flying was the rule and the Cessna 182 was the aircraft at most DZ's, exit separation wasn't too much of a big deal - you gave the other group (if there was another group) some time and then you went. With the aircraft in popular use 15-20 years ago, it was hard to exit very quickly to begin with, and so the issue never came up very often.

Bill von Novak started skydiving in 1991 at a small DZ in New York. Since then he has become an S+TA, an AFF, tandem and static line instructor, and has set two world records in large formation skydiving. He lives with his wife Amy in San Diego.

Since then, several factors have conspired to make exit separation more of an issue. First off, there are more people freeflying. Freeflyers, especially head down groups, drift differently than belly flyers, and thus need different considerations when planning for exit separation. Faster canopies mean that people who open facing each other need more distance to deal with a potential collision. Large aircraft with big doors can hold several larger groups, and those groups can get out those big doors more quickly. Finally, GPS spotting has removed some of the delay between groups. It's rare to see people even check the spot before beginning their jam-up.

I first became aware of this issue in 1994, when I started jumping at Brown Field in San Diego. We went through a series of aircraft as we grew, from Cessna 206's to King Airs to Beech-99's, none of which had GPS. In addition, we were less than a mile from the US-Mexico border, which meant our jump runs had to be east-west and our spots had to be dead on. Several instructors were "designated spotters" and we would argue over 100 yard differences in jump run offset and exit location. After a while we got pretty good at spotting.

As our aircraft became larger, exit separation became more of an issue. We had a few close calls, and so we agreed to start allowing more space between groups. At first it was essentially trial and error - we would leave some amount of time (10 seconds or so) between groups and increase that time whenever someone felt they were too close to someone else. After a while, we began to get a feel for how much time was required. We knew that if the upper winds were strong and the plane was just creeping along the ground, we had to leave more time. We also knew that if we let the freeflyers get out first, we had a problem almost every time. We ended up with a system that worked for us, and had essentially no problems with collisions or close calls after that.

During this time I was also traveling in the summers to different boogies and I noticed a wide variety of exit separation techniques. By far the most common technique was some amount of fixed time - the next group would pause, then climb out and go, without knowing what the upper winds were doing or what the spot was. The next most common technique was similar but they added a "leave more time if it's windy" clause to their delay. There was also a class of jumpers who looked out the door to tell how much separation to leave; these jumpers either looked at angle of the departing group or the ground to tell how much space to leave.

This got me thinking. What really works and what doesn't? I tried a few methods on my own, from the "45 degree" method to a purely ground-based method. After some experiments, a group of skydivers collaborated via email and internet and came up with the actual math behind separation, the physics that determines how far the center of group A will be from the center of group B after they open. But before diving into the math, there are a few basic concepts to cover.

1. What we care about. When we're talking about separation at opening time, we don't really care about where we are in relationship to the plane or even the ground - what we care about is how far we will be in the air horizontally from the next group that opens. So for our purposes, the airplane and the ground don't really matter, and someone watching from either of those places may not get the same "picture" of things that we get. (Of course, we do care about our relationship to the ground when it comes to spotting and landing on the DZ, but that's a separate issue.)
   
   
2. How we fall. In most freefall (tracking dives and wingsuits excepted) we fall essentially straight down with respect to the air. If there's wind, the wind blows us at whatever speed it's blowing. If the wind is doing 30kts at altitude, a group of skydivers will be doing 30kts as they drift with the wind. It's also important to realize how your trajectory changes after you open. At a freefall speed of 100kts, a 30kt wind will slightly deflect your trajectory, because it's a small fraction of your total speed. Once under canopy and descending at 10kts, it will deflect your trajectory a tremendous amount, since it is now a very large part of your speed. Of course, under canopy you have much more control over your own horizontal speed, and the winds may add or subtract from your canopy's groundspeed depending on the direction you are facing.
   
   
3. Speeds. When discussing speeds, it's important to define units. There is feet per second, which is very useful for people who are trying to figure out how far they want to be from another group. At 100 feet per second, 10 seconds gives you 1000 feet, which is about as easy as it gets. You may also hear the terms indicated airspeed, true airspeed, and groundspeed, in both knots and miles an hour. These can all be converted back and forth as needed .

Now that all that's out of the way, the math is pretty simple. The distance you will get between group centers is the speed of the aircraft plus the speed of the winds at opening altitude, multiplied by the time you leave between groups. That's it. So if the aircraft is flying into the wind doing 80 knots per its GPS, and the winds at opening altitude are 10 knots from the same direction, and you are waiting 10 seconds between groups, you are going to get (80+10 = 90 kts, which is 153 feet per second) 1530 feet between groups.

It gets a little more complicated when the winds are not from the same directions. If the winds at opening altitude are opposite jump run, you have to subtract them rather than add them. If the winds at opening altitude are from the side, it's the same as zero winds at opening altitude when it comes to separation.

If you put these equations into a spreadsheet and play with the numbers, some basic patterns emerge. If the headwinds at altitude are strong you have to leave more time. If the plane is slow (i.e. it's indicated airspeed on jump run is low) you have to leave more time. If the winds at opening altitude are strong as well, and from the same direction, you can safely leave less time. (Or, preferably, just leave the same amount of time and you'll end up with even more separation.) If the winds at opening altitude are opposite from jump run, that's the worst case, and you have to leave even more time.

Some people have a problem visualizing how winds at opening altitude can possibly cause them trouble if they leave enough distance on exit. The question is usually phrased as "don't all jumpers follow the same path out of the plane?" And they definitely do. To visualize why this can still cause you problems, take a look at the separation diagram shown below.

Drawing showing exit separations

In the first drawing, there is no wind after exit, and the first group breaks off, tracks, opens, and flies their canopies away from the center for the first few seconds, which is what they should be doing on most formation skydives. (After that, it's a good idea to turn away from line of flight once you're sure you are clear of others in your group.) The second group arrives 10-15 seconds later, shortly after the first group has opened their parachutes, with some room to spare.

The second drawing shows what happens when there are winds are the same all the way down. Notice that the "cone" caused by the breakoff and the canopy flight has shifted strongly to the right. This is because (as mentioned before) once their parachutes are open, the wind affects their trajectory more strongly. As with the first example, it is assumed that everyone flies away from the center for the first few moments. That means the jumper flying into the wind makes no progress and comes straight down, while the jumper flying downwind gets a boost in groundspeed..

The third drawing shows where you can run in to problems. In this drawing, the winds after exit are from the opposite direction. You get the same skewing of the cone, but now the edge of the cone is getting dangerously close to the trajectory of the next group. This is a case where the same separation at exit led to trouble because of opposite winds at opening altitude.

This leads naturally to the question "how much separation do you really need?" That depends on the group. 1000 feet should probably be an absolute minimum for any belly formation skydiving. That means that two four-ways can exit, fall straight down the pipe, track 300 feet from center on breakoff, and then still have 300 feet to deal with avoiding a potential collision after opening. With the speeds of today's canopies, that's a bare minimum. If the group size grows to two 10-ways, 2000 feet might be a wiser separation. If a low-time RW group backslides a bit, again, 1500 feet might be needed to be clear of them at opening time.

So how does a jumper who doesn't want to carry around a calculator figure out how much time to leave between groups? One very simple way is to just look out of the plane and wait until it has covered 1000 feet, then go. This method, originally suggested by Skratch Garrison, takes much of the figuring out of exit separation. It can be hard to determine how far 1000 feet is on the ground, but fortunately most DZ's come with a handy ruler - a runway. A 3000 foot runway allows you to put 3 groups out along its length with a bit of margin thrown in. This method also has the tremendous advantage that it requires people to look out the door, and that means they are more likely to see traffic, high canopies or clouds that could pose a hazard to their skydive.

Another simple way is time-based. There are several tricks you can use to determine how long to wait. One common one is to always leave at least 7 seconds, then if the upper winds are strong divide them by 2 and wait that number of seconds. (Faster aircraft sometimes use divide by 3.) So if the winds are 30kts you wait 15 seconds between groups. This technique uses some math but isn't too bad.

A third technique that seems to be popular for some reason is the 45 degree method. In this method, jumpers wait until the previous group passes through an imaginary 45 degree line before they exit. The problem with this method is that the jumpers never pass through that 45 degree angle, or pass through it so quickly (under 1 second) that it's not useful for determining separation. The numbers confirm this. What you see out the door depends purely on speed of the aircraft, fallrate of the jumpers and type of exit. If the plane is going slower than freefall speed, the group may start out above the 45 degree line, but will drop below the line in less than a second and never rise above it again. If the plane is going faster than freefall speed (which is rare) the jumpers stay above the line and never cross it at all. A good head-down exit will tend to move jumpers lower in the picture. Winds will not affect the picture; an exit in 5kt uppers looks the same as an exit in 50kt uppers.

There has been some friction over this issue. The 45 degree method has a lot of supporters because it's so simple and makes a sort of intuitive sense. Beyond that, it actually seems to work for some people - although it's likely that the extra time it takes to locate and stare at the previous group has something to do with the reason the next group usually leaves enough time. To show that this doesn't work, two cameras were fixed at a 45 degree angle and mounted on a boom outside an Otter's door (see pictures below.) Pictures and video of several jump runs both into the wind and downwind were taken and magnified to determine how close each group was to the imaginary 45 degree line, which was essentially the center of the images. The pictures confirmed the basic problems of the 45 degree rule. RW groups, falling a little faster than the aircraft, never quite passed behind the 45 degree line. Freeflyers, going much faster than the aircraft, stayed well below the 45 degree line for as long as they were visible in the stills (about 30 seconds.)

Some version of the 45 degree method may work for some people. It may be that the simple act of looking out the door delays them enough, or their subconscious may see the group moving slowly along the ground (because the aircraft's groundspeed is low) and send a warning message to the rest of their brain - "hey, hold up a minute." But waiting for a true 45 degree angle simply does not work.

Another issue that has become more important lately is exit order. Some places still put freeflyers out first, and that doesn't make much sense. In 30kt uppers, a belly flyer who leaves 10 seconds and gets out after freeflyer will open 100 feet from him, but if the belly flyer goes first and the freeflyer leaves the same time he will open 2200 feet from the freeflyer. RW groups, since they are in freefall longer, drift farther downwind before opening. It seems like a no-brainer to choose an exit order that used this to your advantage and increased, rather than decreased, separation distances. You can certainly wait 20 seconds after the freefly groups before the belly groups exit if there is some other reason why the freeflyers have to exit first, but at most DZ's it's hard to ensure that 20 seconds, especially since waiting so long almost guarantees long spots or a goaround.

Below are two diagrams that show how exit order can affect separation.

One reason given at DZ's to explain a backwards exit order is that freeflyers open sooner and therefore are beginning to descend before the next group gets there. Bryan Burke of Skydive Arizona has pointed out that you simply cannot trust vertical separation - one premature deployment or malfunction and all that vertical separation is gone. Even during a normal skydive, when you add up altimeter error, pull timing and snivel distance, you can easily get a jumper opening 1000 feet from where he expected to be open. In fact, Bryan points out that at Skydive Arizona, the primary reason high pullers get out last is not for separation but rather because they are the ones that can make it back from a bad spot.

Every drop zone is going to have a different set of rules and a different approach to exit order. Some work well, some don't work as well. Jumpers have to understand the factors that can reduce group separation so they can make informed decisions about when they want to exit and what kind of exit orders they are comfortable with.