Driving down a narrow twisting country road in the Dordogne region of France most tourists stop their cars to look at the idyllic roadside pool overlooked by a classical old French farmhouse and watermill. Quietly signed as the 'Source de Coly', staring into the clear blue water it's just possible to make out the small dark archway from which the underground water wells up into the daylight. The submerged cave system that stretches for many kilometres from this tranquil spot has over the past ten years seen some of the greatest advances in underwater cave exploration both in terms of distance explored and the equipment & techniques used to accomplish these dives.
The Doux de Coly cave was first dived in the 1960's, but it was not until 1970 over a series of five dives that P J Debras made any significant penetration reaching a point 365m into the cave, at a maximum depth of -52m.

In the early 1980's a Swiss team, Groupe Lemanique de Plonger Souterrain (GLPS), turned their attentions to the site. This highly motivated team were bringing a new approach to European cave diving by using underwater scooters and stage diving techniques to make significant progress in many of mainland Europe's classic cave-diving sites. In five days in August 1981, a team of GLPS and French cave divers pushed the Doux de Coly to a point 1760m from the entrance - mostly at depths of around -55m using their underwater scooters. These allowed faster movement through the cave, and the use of less breathing gas than a finning diver.

In 1983 the same team returned, and the cave was pushed to 2100m, continuing at similar depths. The following year, 'siege' tactics were employed, with numerous dives being carried out to allow the gradual establishment of relay dumps (i.e. full gas tanks left at pre-arranged points for the diver to use). These stage dumps at 700m, 1100m & 1500m, were augmented with extra emergency bottle dumps at 550m, 900m & 1300m. This system enabled the group's lead diver, Olivier Isler, to continue with the exploration to an amazing 3100m from the entrance in a dive lasting 3 hours plus a further 8 hours of decompression.
Following this dive Olivier realized that future explorations would not be feasible using ordinary 'open circuit' breathing equipment and began to look at alternative systems. As a result, following the 1984 expedition, Olivier and his friend Alain 'Ron-Ron' Ronjat (a French Naval Technician) set about developing a semi-closed re-breather system. This would recycle the diver's exhaled breath - removing the carbon dioxide and topping up as necessary with new breathing gas. This would make his gas supplies last five times longer than on open-circuit gear. The RI 2000, as it became known, was to prove the key to future explorations at this and other similar sites.

In the summer of 1989 Olivier and his team returned to the Doux De Coly. They made no real progress on this occasion spending most of their time sorting out teething problems with the new rebreather system. Much of the work involved streamlining the unit to enable it to be used at speed with the Aquazepp scooters. No further progress into the cave was achieved but many lessons were learned which would prove invaluable later on.

During this same period, a small number of members of the British Cave Diving Group (CDG) had also been visiting the Dordogne area of France for a number of years. The warm, clear water and large cave passages contrasted sharply with the small, cold, murky sumps found back in Britain. On one such trip John Cordingley had bumped into Olivier and his team, and as a result John, Russell Carter and Malc Foyle became involved in Olivier's explorations at the Emergence de Ressell in the Cele Valley. This in turn led to a number of CDG members - John, Russell, & Malc together with Robin Brown, Brian 'Scoff' Schofield, Brian Smith & Carol Tapley - being invited along to participate in Olivier's 1991 Doux de Coly expedition. They were to join with a twenty strong team of divers from Switzerland, France and Belgium.

Because of the technical difficulties and the specialized equipment used, Olivier was the only diver able to push on at the end. The nature of the expedition was thus a little unusual in that it involved many man-hours being expended in order for one man to explore, alone, the end of the known cave. The effect of this was two-fold - it left Olivier under enormous pressure to succeed, whilst the other divers knew that they were little more than supporters. Nevertheless, Olivier was the first to acknowledge the role that the rest of his team played - likening the expedition to a space mission where everyone played a vital part, without which the end result could not be achieved.

This 1991 expedition was a great success. On his first push, Olivier reached a point 3800m into the sump, of which a significant part still lay below fifty metres depth. A few days later a second push was mounted during which he reached a point 4055m from the entrance, smashing the world record for the longest single sump. On this dive Olivier spent some 14 hours underwater. And the cave still continued...

The RI 2000 and twinned Aquazepps proved themselves on the 1991 expedition but possibly the biggest single technological advance on this dive was Olivier's decision to use an underwater habitat, to make his nine hours of decompression safer and more comfortable. This had been done previously in Florida cave entrances but the nature of the Doux de Coly meant that the habitat had to be placed a long distance inside the cave system, something which had never been attempted before. The habitat was a large inverted metal tank with a ballast bin attached below. Fitted with a fold down seat, it formed a small airbell, which allowed the diver to get his chest and head out of the water. It was placed some 300m from the entrance, in the shaft that descended vertically from -3m to -43m. Attached by a complex rope system, the habitat's buoyancy meant it could be raised in stages up the shaft, as Olivier's decompression progressed. This combination of a relatively comfortable decompression facility, the Aquazepp scooters and the RI 2000 rebreather enabled Olivier to push forward the boundaries of cave exploration still further, but things were not to finish here - the cave continued...

In 1997 Olivier returned to the Doux De Coly but problems with the weather and an accident with a scooter 2000 metres into the cave which resulted in serious damage to the rebreather meant that no further progress was made on this occasion. The cave was however extensively relined up to the 1500 metre mark in preparation for another attempt at a later date. That attempt eventually came to fruition earlier this year when a strong team of 10 British CDG divers joined Olivier and 10 other Swiss and French divers in the South of France. The first week of the project was spent on preparation. The main shaft 300 metres in had to be rigged with decompression lines, safety cylinders and the two decompression habitats that were to be used.

As well as Olivier, the German diver, Reinhard Buchaly was also to make some major penetration dives in support of Olivier's attempt on the end. Reinhard's approach to the dive was classic WKPP style as adopted in Florida on their recent record breaking dives at Wakulla Springs. Rienhard was to dive using a Halcyon rebreather and open circuit bailout gas. Using this system effectively as a solo diver unlike the WKPP divers who always dive as part of a team, meant that he was restricted by the amount of bailout gas he could stage deep into the cave and would not be able to reach the end. He was however able to assist Olivier greatly in the early stages of the project, diving with him to the 3km mark, replacing sections of damaged and broken line. Interestingly Reinhard chooses to dive on Trimix using a fixed depth decompression habitat whereas Olivier uses Heliox and a variable depth habitat. On their joint dive to 3000m Rienhard left the water a full 2 hours before Olivier although Olivier was able to do virtually all his decompression in the relative comfort of the habitat while Rienhard had to do over 2 hours in the water before he could enter his habitat. Part of the reason for this is that on his big pushing dives Olivier chooses to run a saturation table which gives him a predetermined 9.5 hour decompression virtually regardless of bottom time once he goes over a certain runtime. Whilst this gives a very long decompression it means that he has a very predict- able schedule and very few restrictions on the limits of exploration. The only alternative would be to have to run multiple dive profiles for different scenarios as the nature of original cave exploration like this is that no one knows what the cave may choose to do.

Both divers used heated undersuit systems to combat the 12-degree water temperature, plugging into external battery systems on returning to the shaft. The new habitats consisted of a neoprene chamber supported by a ballistic nylon shell and an aluminum framework. Rienhard's was attached directly to the walls of the shaft using rock anchors whilst Olivier's, which was larger, was connected via a tirfor winch to cables and ropes anchored to a large car sized boulder at the base of the shaft. Olivier is unable to easily remove the RI 2000 unit underwater and wears it for the entire duration of his dives. External cylinders of decompression gas are plugged directly into the unit thereby maintaining the advantage of the rebreather system even through the decompression phase of the dive. Rienhard removes his Halcyon unit prior to entering his habitat and does the decompression from then on using open circuit gas. This he admits is the most dangerous part of the dive for him. Removing all your life support systems whilst hanging part way down a 50m deep shaft, 300 metres inside a cave has to rate as pretty dangerous in anyone's book! The main risk is in dropping something (like the rebreather), or losing control of buoyancy whilst removing such a heavy piece of equipment. For this reason two minders would always be with him whenever he entered the habitat or left it. In practice whenever either of the two lead divers were doing long decompressions or using the habitats at least one support diver would also be on standby in the shaft. Some of these babysitting stints would last up to three hours. Doing long decompressions is bad enough but it felt very strange to be doing the decompression stops for someone else's dive!

After almost two weeks work rigging the cave, replacing and repairing lines and installing and testing the habitats Olivier was finally ready for the big dive. Early in the morning two huge U.S. style scooters similar to those designed by Bill Gavin and used by the WKPP were swum into the cave and placed at the deep point at a depth of 60m, 350m from the entrance, a short way beyond the shaft. The habitat was also moved to its deepest decompression depth at 18 metres.

During the morning Olivier assembled the RI 2000 unit at the waters edge. A triple redundant rebreather system, it incorporates 2 x 20L cylinders and 2 x 12.5L into the rig itself and via high pressure connectors can be supplied with an infinite amount of gas from external tanks. The back-mounted unit contains two independent rebreathers and the cylinders whilst the front-mounted ventral unit gives the third independent system. Any of these 3 systems combined with any one of the four cylinders carried, has more than enough capacity to get the diver home from the furthest point of penetration. Breathing gas used is Heliox 25 in all four cylinders with Nitrox and pure Oxygen plugged in as required for decompression. The max PPO2 is maintained between 1.2 & 1.6. Unlike most rebreathers the unit measures the diver's work rate by a system of electromagnetic valves which measure the volume of gas the diver is breathing through the loop, adding extra gas as required. Each system uses twin injectors in case of failures.

The breathing bags are inside the unit for protection and mimic exactly the diver's own lungs to give minimum breathing resistance. Each unit uses a 3.5L scrubber canister (split into two on the ventral unit for a lower profile) giving a minimum of 10 hours run time on each of the three units. Designed specifically for Olivier's cave exploration projects over a period of 5 years it took over 5000 hours to design and build and weighs a colossal 110kg in air! To date Olivier's deepest dive using the RI 2000 has been -155m but with a glint in his eye he told me he has plans...

At 1:00pm Olivier left the sunshine and the assembled onlookers surrounding the Doux de Coly sump pool and supported by one other diver headed off into the cave. Dumping his small transit scooter at the shaft head he dropped into the depths below, picked up the previously staged primary and backup scooters and headed off into darkness alone. A small Hyball ROV (remote operated vehicle) had been previously positioned in the shaft at -35m to await his return and monitor the decompression phase of the dive. After 4 hours of watching an empty shaft the tension started to build, as did the number of team members gathered around the small surface video monitor relaying pictures from the ROV. Even though everybody knew that the expected bottom time was going to be anything from 4 to 6 hrs (although rumours abounded amongst the team that Olivier had the capacity for up to 8 hours and might just use it) things started to get quite tense after 5 hours. The tension evaporated when 5 hours and forty minutes after he left the shaft bubbles and lights on the video monitor signaled Olivier's return.

Support divers immediately entered the water to check that all was well and found Olivier happily hanging in the shaft connected to his decompression cylinders and no doubt contemplating the 9.5 hour schedule ahead. All had apparently gone well although at this stage he was not letting on how far he had gone. I took the opportunity to snap some quick shots as he made the short deeper stops and with little else to do and nowhere to go he posed obligingly. After 1:15 mins he made the transfer into the habitat where finally after nearly 7 hours underwater he was able to momentarily remove his mouthpiece, take a drink and speak to the surface using the habitat's telephone system. The habitat maintains an air atmosphere at all times in case of an Oxygen convulsion during the decompression so the relief of removing the mouthpiece was short lived as all the remaining decompression had to be done using the RI 2000.

The next 8 hours proved traumatic for Olivier who described them as the worst hours he had spent underwater. Too tired to operate the very stiff tirfor winch which raised the habitat up the shaft he had to rely on support divers to make the moves for him and numerous small problems occurred with hose routings and location of the deco cylinders. The most dangerous manoeuver came however right at the end as he attempted to leave the habitat. He had previously passed out a list of instructions for how he wanted to make this move but it didn't help much as they were written in French and neither Dave Ryall nor myself who were on babysitting duty at the time understand much written French! After passing back in another slate asking for instructions in English the whole page of French suddenly got condensed into two words - 15 mins! At this stage my time was up and I left for the surface ready to photograph the triumphant return, leaving the departure procedure in the capable hands of Dave and Russell Carter who had arrived to relieve me.

The problems with hoses and deco cylinders compounded themselves as Olivier left the habitat and at one point he was left with no other choice than to totally disconnect from everything, leave the habitat and then reconnect to his final deco mix. Totally exhausted after 15 hours in the water he had to be helped across to the top of the shaft and the waiting scooter ride back to the entrance.

Escorted by the support divers he eventually emerged into the glare of video lights and camera flashes at 4:30am, cold, and totally exhausted, but successful after a dive lasting 15.5 hours. From his previous end point in a blind chamber at 4km from the entrance he had relocated the way on at 3.9 km and had pushed the cave to 4.25km - a total dive distance of 8.5 km! The bottom time was 5 hour 40mins at a maximum depth of 60 metres, with a further 9 hours 30 mins decompression. The RI 2000 though now nearly ten years old proved itself yet again as the ultimate cave diving exploration tool. For the entire bottom time of over 5 1/2 hours the unit only required 5.1 cubic meters of gas, which equates to just thirds on the integral 20L cylinders. The two 12L cylinders were left untouched!

It wasn't the longest cave dive ever - that record rests firmly with the WKPP project diving at Wakulla Springs in Florida who have now touched the 6km mark. But all caves are different and Olivier's dive is certainly on a par with those being done in the U.S. Permission to dive the Doux de Coly is not readily given with access granted only every couple of years in return for large sums of money making familiarization with the site and the continuity of the exploration very difficult. The water temperature is 12 degrees compared to rather more diver friendly 22 degrees in Florida and probably most importantly, both logistically and psychologically, due to the nature of the cave the decompression has to be done inside the cave, a long way from the entrance. Equipment and technology are important at this level of cave exploration no doubt, but equipment is available to all and is only a tool. Attitude, psychology and the right mindset are far more important than any number of toys and in this Olivier is surely a breed apart.

By Gavin Newman.