The Dragon’s Breath: Bailed DGT April 2019

Some background

The Drakensberg Grand Traverse (or DGT for short) is nothing new. I don’t know who recorded the first variation of the route, but I believe it was at least 40 years ago. There have been many different variations in direction and line, but the concept is roughly that of following the border between KZN and Lesotho from one end to another.

Records for the DGT have been kept to a very limited extent. The table of who holds the fastest times and the rules for a speed DGT are well maintained, but statistics such as the oldest, youngest and most completed DGTs are very limited and generally assume that we would know of any more impressive feat.

My personal views on a DGT

I don’t remember how I first heard of the DGT, I just always remember it being on my to-do list. Even before I started hiking. Every traverse I have done is written up online in detail, so I won’t rehash my entire history here – but in short, in January 2017 I completed my 6th DGT, two of which qualify as speed traverses and feature on the list of fastest DGT times. At the time I was on a perfect score, six starts and six finishes.

My traverses were made up of three in each direction. There had been some rough patches along the way – hypothermia was included in two of the stories, there were some massive thunderstorms, gunshots, stolen gloves, running out of food and water, missing resupply packs, dogs and a lot more – but we always finished. Those six traverses covered 55 days of hiking and 153 summits (including duplicates). Only two of them actually finished in the planned number of days.

I believe that always hitting your goals on the first try means that your goals aren’t hard enough. But goals can be complex – and finishing a DGT in less than 96 hours (4 days) has been a goal of mine for four years now. And a summary of how it has gone is as follows:
1) In November 2015 I set off with Andrew Porter for a four day DGT and ended up finishing in 107h51.
2) On 31 December 2016 I set off with Andrew Porter and Mike van Wyngaard and finished in 126h55 after getting caught in a massive storm. We spent a day and a half at Sani waiting out a storm, so it didn’t even qualify as a speed traverse as it was not unsupported.
3) In April 2018 I set off again with Mike van Wyngaard and unfortunately this would be the time I finally learned that starting a DGT didn’t guarantee finishing one. I knew something was wrong when we started ascending Stimela Ridge just over 20km into the traverse and Mike needed to stop multiple times. By the second night we were barely past Cleft Peak and after Mike spent the entire night coughing, we went down Thuthumi Pass and my perfect run of completing every DGT that I had started was over.

When you keep missing a goal, you have to ask some hard questions:
– Why do I keep missing this?
– Am I training hard enough?
– Is my strategy good?
– Do I have the right team mate?

So when Warren and myself set a date to have a go at a 4 day traverse, I knew I had to up my game. We had been talking about doing this for years, and in the lead up to the traverse, I did a 117km solo hike of the Giants Cup Trail back-to-back in 36h14, as well as various other difficult hikes. I also focused more on my running to improve my speed. I managed to improve my Park Run times by almost 4 minutes over two years. I also started doing a regular 5km loop with 230m in altitude gain and loss a few times a week to up my power on hills.

Along with speed training, I managed to reduce my pack weight by 1.6kg – meaning that I was about to start a 4 day DGT with a 28 litre pack.

We picked our timing for the traverse to be over the Easter weekend – thus combining a long weekend and full moon. Our strategy was simple – big breakfast and an 8am start. Day one would be short, with Easter Cave as a goal. Day two would be long and hard, pushing to Upper Injisuthi Cave. Day three would be the most distance and the least vertical of the days, pushing to just past Thabana Ntlenyana. And from there we would push to the finish on the fourth day, with the option of finishing up to 8am on the fifth day if required.

Day 1

So after a lovely breakfast at Witsieshoek, Warren said goodbye to his wife and kids and we jumped into the shuttle for a trip up to Sentinel Car Park. The road is in very bad shape right now.

After sorting out a few basic requirements, such as putting on sunscreen, we were ready to set off. Our start time was 8:14am


We made a fast start, opting for the high approach rather than the standard trail into the chain ladders. This approach skips the more eroded sections on the walk in, but is on a very well worn path. Probably not faster, but I prefer this line.



We made our way from there to the top of Mont-Aux-Sources in good time, bagging the first checkpoint 2h01 into the attempt. The summit was very windy.


We made our way down to the Senqu and took our first break at the rock pool just below where you begin the climb up Stimela Ridge. It was too cold to take any advantage of the pool, but we knew we wouldn’t see water again for a while, so we filled up and had something to eat.


The wind was howling as we made our way across the plateau, before dropping steeply down to the junction of the rivers behind Rwanqa and Mbundini Pass. The kraals in the area were vacant.

We saw a group of three hikers at the junction. They were on day three of a Northern High Traverse. They were very impressed that we had covered so much ground so quickly. It occurred to me that we were 5 hours into our traverse, while they were more than 50 hours into theirs. But that is the nature of hiking – there are many ways to enjoy it, and what I enjoy most isn’t the same for everyone else.

We made our way up to the cutback highway, stopping for water at the last spot before we left the final side stream. Our pace was good, but I could feel my energy waning. This is to be expected on a big day, but there was still one major hill to go – so I knew I needed to pace cleverly if I wanted a good night sleep.

The cutback highway section was exceptionally windy. We both had gloves and jackets on. As always, this section proves to be fairly long, with some great views mixed in to keep you motivated.


When we eventually dropped down towards the Senqu, we once again stopped on the way down for a break. It would be getting dark soon, and taking breaks in the dark isn’t really ideal.

The long slog up the Nguza valley went surprisingly well. The river was full and the waterfalls kept the views interesting.

Hiking by moonlight is actually awesome. By the top of the valley, we both had our headlamps on, but the moon was eliminating our surrounds enough that the GPS was only useful for confirmation that our line was right.

Heading up Ntonjelana Ridge was slow. As my energy continued to fade throughout the day, my pace dropped too. I could see Warren’s headlamp much higher up the valley, and I was having to stop every few minutes to catch my breath.

The summit of the ridge providing a spectacular view of the Cathedral Peak ridge by moonlight. Naturally it won’t come out in a photo – but it was an incredible sight.

Now that the ground was easier – we were on a trail on flat and downhill, my pace was much faster. We made our way down through Ntonjelana Gap and were soon slogging up the hill to Easter Cave. I really struggled with this last bit, but soon was in my sleeping bag enjoying a nice hot cup of soup.

Day 2

The next morning we were walking by 4:05am. The moon was still high in the sky, which helps a lot. We had a few difficulties finding the trail when it crossed the river and side streams, but for the most part, we made our way up Mahout easily enough.

It made me a bit sad to look at my favourite section of the Drakensberg at first light. I could see Cockade and Elephant, but just the outline. The dramatic cliffs were too dark to clearly make out. By the time we reached the top of Cockade Pass, there was enough light to see some of it – which was great.

We stopped for our first break of the day across from False Tseke Pass. Cleft Peak is a slog and there is often no water behind Organ Pipes Pass, so we filled up here.

Cleft Peak took an age to climb, and the strong wind did not help. While we made our way up, one thing was playing on my mind. I know these conditions – these are Berg winds – which means there is a coastal low. And a coastal low means a cold front will be hitting soon. This was not good news. The Northern Drakensberg is mostly in a rain shadow as a result of the shape of the range – meaning that the worst of a cold front is usually skipped there. But we were about to leave this zone and move into Giants Castle, where the brunt of cold fronts would be felt.


We reached the summit of Cleft at 7:58am, got a summit shot and moved off immediately. It was very unpleasant in the wind. Just under 4 hours from Easter Cave to the summit of Cleft isn’t terrible, but isn’t quite as fast as we had hoped.


The pain of this section of a DGT is that you have to ascent 400m to get over Mahout Ridge, followed by another 300m for Cleft and another 200m for Ndumeni Dome – all in quick succession. Relatively speaking, Ndumeni Dome is one of the easiest ridges of the traverse (north to south, that is) – but to have it immediately after the other two, it takes its toll.

We stopped for a break at a rock pool behind Little Saddle before continuing down the river. The kraals here weren’t empty, but the shepherds kept their dogs under control, so we were fine.

When the valley ends, we were confronted with a choice: risk the dogs and boring valleys of the Yodeler’s Cascades or climb the very steep Didima Ridge that affords amazing views. Both require almost 800m in altitude gain, making it the third largest climb of the traverse, and by far the hardest. Both have the same vertical, but the cascades are roughly twice as long.

We opted for the steeper option. This is where my 2016/2017 DGT attempt had begun to fall apart, so I was hoping to get past it without as much difficulty this time.


The first section went easily enough. But the problem is that this line never ends. It is one of those hills where you just keep going up. Eventually we hit the steepest section, and by the time I reached the top of it, I was destroyed.


We had a break at the top of the gully, still 150m vertical from the top of the ridge. I had hoped I would be able to make up some time on the trail above it, but was so tired from the climb that I took multiple breaks before we reached the top of the ridge.

On reaching the top of the ridge, we then struggled to find the trail, even though we had a GPS track of where it was and had both done it multiple times before. In retrospect, I think I would take the trail to Didima Cave and then follow the ridge up from there – this line is easier and means you can refill your water bottles.

When we eventually found the trail near the back of the U-Bend, I said to Warren that I doubt we will reach Upper Injisuthi Cave before midnight. Once we leave this ridge, there are no caves till then, so we would be fully committed. But if we finish after midnight, we won’t be able to do the planned 64km day three, and due to there being no caves for the last 30km of that day, we would end up at Giants Cave. So we may as well call it a night at Reido Cave and just try to set up for an early finish on day five.

The next concern was that we had no water. While everything had been very wet, we were on top of a ridge and not near any streams. Finding water at 3200m is generally difficult when you are on the slopes of a 3337m peak.

Luckily my GPS co-ordinate for Reido Cave was slightly wrong and lead us to a drip that we used to fill our bottles. It took half-an-hour to collect the 2.5 litres of water we collected, but it was better than slogging down into the valley for it.

While we were sitting here, Warren had begun to cough. And it wasn’t a usual cold weather and altitude cough like everyone gets up there. This sounded bad.

We were in the cave by 5pm, and Warren was not sounding good. I know that exercise with the flu can lead to heart failure, and this is even worse at altitude as your heart is already working harder to deal with the lower oxygen content. I told him that it was his call.

Warren got no sleep that night. And trying to sleep when someone is coughing all night, admittedly I didn’t get much either. But the views towards Winterton with the lights of civilisation and the moon lighting up the valley were amazing to see.

Day 3

I woke up to a cold misty morning. Just before 7am we were packed up an ready to go. Not to Giants Cave as previously planned, but down Grays Pass. The traverse was over. Warren was too sick to continue and had messaged his wife to pick us up at Monks Cowl campsite later in the day.

We made our way to the Didima Highway trail and were soon plodding along towards Grays Pass. Warren needed to stop for regular breaks – which I was quite happy about, I knew he needed to keep his heart rate down.

It was cold and windy on the trail. The mist was coming in from Lesotho, the cold front I had worried about the day before was here. Now to navigate to Grays Pass in the mist with only a GPS co-ordinate and no track. Luckily I knew the trail we were on would lead straight to the pass, so we just followed it.

Unfortunately the trail kept climbing, which was not what Warren needed. He was coughing a lot, and I knew this was not good. Nkosasana Cave would be our last possible stop to stay the night, and no helicopter rescue was going to happen in mist and wind – so our choices would be to wait in a cave until he felt strong enough to head down; not ideal in cold conditions, or to keep going.

The side valleys never ended, and even though I knew exactly where the trail we were on was going, I had never used this section before – so I just had to trust that my route knowledge was solid and that this line would get us there.


Occasionally the weather would clear providing a bit of a view, but the mist had generally been very thick. Then suddenly we crossed a ridge into the Nkosasana Valley and it was completely clear, Grays Pass was right in front of us.


We saw a large group at the top of the pass, they were heading for Ships Prow.

We made slow progress down the pass, enjoying the views and vultures as we went.




When we were down the pass, I knew that a horse rescue would be possible if necessary. The lower section of the pass and the walk out was full of people – unsurprisingly for a long weekend.


We stopped for a few breaks on the way out, and eventually found ourselves at the mountain register at Monks Cowl at 2:42pm, meaning that we had completed our 104km mini-traverse in 54h29, no rescue or outside support required. We had ascended roughly 4400m and descended around 5400m (as Monk’s Cowl is more than 1km lower in altitude than Sentinel Car Park).


Naturally we hadn’t achieved what we set out to achieve, but as we walked towards the hot showers at the camp in rainy and windy conditions, we weren’t sad to be off the mountain. As it continued to rain on Sunday and Monday after the hike, we knew this had been the right call – even if Warren hadn’t had flu, it would have been miserable, and probably dangerous to be up there.

From a personal perspective – it would be easy for me to say that Warren’s flu was the reason we didn’t get this. But the reality is that I was struggling. Realistically I don’t think sub-96 hours would have happened in any event, even if the weather had been better. Sometimes you do a lot to get ready for something, but it wasn’t enough. So I will have to return to the drawing board and train harder. Sometimes in life you have to decide if you really want something, and if so, are you willing to work for it.

Special thanks to Ruth for handling logistics, to my mother for the lift up to Harrismith and to Hi-Tec for the Condor shoes that I used.


  1. Awesome read! I always enjoy your writing, and the photos. Thanks for sharing this. Good luck for the next mission.

  2. Ketja Nhlapo · · Reply

    Interesting write up indeed. Maybe one day, one would experience the DGT as well. Just a bit of clarity though; could your accent and decent have been the 4000m and 5400m, taking into consideration that the highest point is 3482m asl?

    1. A DGT is mostly a question of crossing over ridges, so there is a lot of up and down. The ridges range from 200m to 800m. We crossed 8 ridges before bailing. A standard north to south DGT includes 8800m of total ascent and a south to north requires just short of 10 000m.

      1. Ketja Nhlapo · ·

        Ah okay, thanks for the explanation.

  3. Great read – thanks for the write up.

  4. Stu Summerfield · · Reply

    Thanks for the read. I’m envious of what you both achieved. Sorry you guys didn’t achieve your primary goal but glad you made it home safely. When you have a wife/kids waiting at home, there’s always that added weight of responsibility in the back of your mind (well at least for me) and if you get sick it’s difficult to make that call to abandon the challenge or keep pushing through. Still on the hunt for my own DGT once I get back to KZN again.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: