Your GPS Can Kill You

(required reading for every sailor who uses a GPS)

It all started normally at the Cruising Yacht Club of Australia in Sydney on the evening of Friday 9th October 2009.

The ocean racing crews were busy setting up their boats for the short (92nm) overnight ocean race to Flinders Islet about 40 nm south of Sydney and home again. The weather was 25-30 knots from the south and the seas lumpy but not considered a problem. To many this was only a short sprint race and they expected to be home for breakfast.

One of the favourites for the race was the 9 year old maxi yacht Pricewaterhouse Coopers (PwC), a Reichel Pugh 80. The owner Andrew Short was a very good and experienced sailor and he had a crew of 18 aboard, most of whom were very experienced and capable sailors who had campaigned the boat for the previous two years.

The race started at 8pm. Six hours later PwC was a total loss on the rocks at Flinders Islet. Andrew and another crew member Sally Gordon were lost and the rest of the crew were lucky to escape the same fate.

How could this have happened to a modern, well maintained, fully crewed racing yacht sailing in its home waters?

just two reasons

Reliance on GPS and chart plotter for coastal navigation and fatigue.

The Cruising Yacht Club of Australia has just released its internal investigation into this disaster and you can read the full story here.

But in a nutshell, this is what happened.

Andrew always sailed as the skipper, principal helmsman, tactician and navigator. On the day in question he had come directly from a full days work and took command of the boat. He then steered, navigated and skippered the boat for the next 7 hours until the boat hit Flinders Islet at approximately 2.35am on 10 October 2009.

In the sea and wind conditions that night, steering and navigating for 7 hours straight was not a good idea.

PwC was equipped with 2 Garmin chart plotters, one of which had a readout on deck near the starboard wheel. The navigation station below was not used and the deck readout was the only information Andrew used to navigate the boat that night.

Flinders Islet is an unlit, low islet just outside Port Kembla south of Sydney. It is notoriously difficult to see at night due to the brilliant haze of shore lights behind it. Before GPS we would all give it a very wide berth in any heavy weather because we were never sure where it was.

As PwC approached Flinders Islet Andrew was comfortable with his position as shown on the deck readout. They were on port tack and it was only as they saw the islet that they realised they had overstood it and were well to the east of it. The boat was eased away to starboard to clear the northern end of the islet. Andrew was still using the GPS, and his forward view was obscured by the spinnaker bag hooked onto the port rail forward.

Close to the islet the bowman felt the boat surge and saw breaking waves ahead. He yelled “Come away, Come away” and the boat came right away to starboard to a heading of about 335 degrees – almost back the way it came.

But it was not enough.

PwC hit the north end of Flinders Islet and was smashed to pieces on the rocks. Andrew, who was the only one not wearing a Personal Flotation Device ( PFD ) was washed overboard and lost. Sally Gordon was also washed overboard and lost. The rest of the crew scrambled ashore onto the islet and were rescued.

the navigation gods blinked

Andrew had used his deck chart plotter to navigate from the start of the race. As he approached Flinders Islet the navigation gods blinked.

At 2.00 am on 10 October 2009 there were 7 satellites visible above the horizon, but only 3 were high enough to be usable. Of those 3, 2 were on courses that differed by less than 20 degrees. Two satellites this close together cannot give two accurate lines of position. So effectively PwC’s GPS was working with 2 satellite positions only.

To this you must add the problems inherent with GPS positioning and yachts. The GPS antennae on a yacht is usually close to the water and the boat is heeling and bouncing in the seaway. This made any position displayed by the chart plotter aboard PwC very suspect indeed.

Interestingly, there was an underwater tunnelling project underway close by that used a very sophisticated GPS system whose accuracy was constantly monitored. They shut down the tunnelling operation at 1.30 am that morning due to unacceptable errors in their GPS positions.

The Dilution of Precision (DOP) graph shows the large increase in dilution (increased error) of the GPS position results between 1500 UTC (2.00am local time) and 1630 UTC (3.30am local time that morning. The error indicated by the DOP graph is for a fixed, land based GPS unit. The error would have been much larger on PwC due to the low antennae, the angle of heel and the yaw and bounce of the boat.

Andrew was tired and he was navigating with a system with an error that could have been up to +/- 100 meters.

As well as the GPS and chart plotter errors, the guaranteed horizontal accuracy of the paper chart at Flinders Islet is about +/- 50 meters. Most of the Australian coast is charted to a guaranteed horizontal accuracy of +/- 500 meters. These parameters are noted on paper charts but only on some electronic charts. This means that what is displayed on the chart plotter needs to be confirmed by other navigational means and not accepted as anywhere near 100% accurate.

what are the lessons?

Lesson 1 . Don’t think you can do everything on a boat. Spread the load and make sure the navigator and skipper get enough rest. Initiate a good watch system and allocate responsibilities before you leave the dock. PwC had no formal watch system for this race, and because it was a relatively short one, it is probable that no-one had any proper rest.

Lesson 2 . NEVER, NEVER, NEVER assume the position you see on any chart plotter, or GPS readout for that matter, is accurate. It’s only as good as the satellites it can see, and as you can see from this disaster, the errors that can be introduced without your knowledge can be fatal.

Lesson 3 . Never assume your charts, whether paper or electronic, are accurate. I’m reminded of the story about the newly qualified navigator who calculated and plotted his position for his skipper to the last 1/10th of a nautical mile. He showed his plot to the old skipper who grunted and placed his open hand over the fix on the chart and said “So you think we’re somewhere around here”

Lesson 4 . Eyeball navigation 101 is still the best when close to the rocky bits on the edge. Followed closely by manual 3 point fixes, bearing and distance off calculations and any fix that uses landmarks and features you can see.

Save the GPS for offshore passages where +/- 100 meters is of no consequence.

Lesson 5 . The new PFD’s are comfortable and they don’t interfere with any work on a yacht. Always wear one when sailing at night, when sailing alone or at any time when the wind and sea conditions dictate.

I knew Andrew and competed against him in many ocean races. Sally Gordon was a well respected racing navigator. Their deaths are a great loss to the ocean racing fraternity in Australia.