$ cat diving/expeditions/lusitania-2006.plan

lusitania.

Two expeditions in the summer of 2006 with the Shadow Divers. In May we spent a fortnight trying to dive her, and the Celtic Sea won every day. In July we went back and made three successful, very respectful visits, with a Megalodon for the bottom and hanging stations for the long way up. Most of them ran past eight hours in the water.

Where · 11.5 nmi south of the Old Head of Kinsale Seabed · 93 m May 2006 · 15 attempts, all abandoned July 2006 · 3 visits Plan · 8 h 2 min runtime Site · RMS Lusitania →

$ cat ~/shadow-divers/crew

01 / The team

Eight divers and a skipper, and everyone with a second job.

We each maintained our own kit. Everything else was shared out.

GASES

Peter

Kept the gases right: every mix, every fill, every cylinder charged.

dived with Kim

SAFETY

Graham

Our safety man: confirmed every gas, planned the dives, and made sure we dived the plan.

dived with Doc

DECO LIBRARY

Kim

Made sure we had reading material for the deco stops, and was my main buddy on every dive.

dived with Peter

THE DOC

Doc

Looked after the human factors, and ran all the photography and video.

dived with Graham

PENETRATION

Barry

Laid the lines for any penetration we tried, and safety-checked the wreck.

PENETRATION

Glen

Laid the lines for any penetration we tried, and safety-checked the wreck.

HISTORIAN & PERMITS

Sal

Our duty historian: dug up everything we ever needed to know, and made sure we had every permit required.

dived with Jonno

STORES & ENGINEERING

Jonno

We each looked after our own kit; Jonno made one hundred per cent sure all of it was in perfect working order.

dived with Sal

BOAT

Simon

A non-diver, who kept everything topside working beautifully.

topside

$ grep lusitania diving/log

02 / May & July 2006

Fifteen days of no, then three of yes.

She lies on her starboard side in 93 metres, eleven and a half nautical miles south of the Old Head of Kinsale.

11–25 MAY 2006

We ran one dive a day. Robbo did make the wreck, but the conditions were dreadful, and every attempt was abandoned.

3–14 JULY 2006

Three successful, very respectful visits. Each dive was a combination of my Megalodon rebreather and open-circuit deco gas from the hanging stations, and most of them ran past eight hours.

U-20 torpedoed Lusitania on 7 May 1915 and she sank in eighteen minutes; about 1,200 people died. She is a war grave, and since 1995 an Underwater Heritage Order has protected her: diving her needs a licence from the Irish State, and nothing may be disturbed or taken. Sal made sure we had every permit we needed. We went to see her, and we left her as we found her.

$ php bin/build-dive-plan.php --print

03 / The plan

Rebuilt, like the log.

The plan sheets went the way of the logbooks. The site, the kit and the runtimes are what we dived; every number below is worked out again from scratch, with the model we'd have used.

RUNTIME

8:02

8 h 2 min in the water

BOTTOM TIME

55 min

at 93 m, from leaving the surface

ON THE LOOP

2:40

down, bottom, and stops to 24 m

OPEN CIRCUIT

5:22

from the stations, 21 m up

0 m

20 m

40 m

60 m

80 m

100 m

0 h

1 h

2 h

3 h

4 h

5 h

6 h

7 h

8 h

ON THE LOOP · 2:40

OPEN CIRCUIT, FROM THE STATIONS · 5:22

93 m · 55 min

surface 8:02

Depth against runtime for the plan. The shaded part is on the loop; the rest is open circuit from the hanging stations. Every stop is in the table below.

The runtime

Depth	Stop	Leave	Breathing
0 → 93 m	4.7 min	0:05	Descent, loop at 0.7
93 m	50 min	0:55	Bottom, loop at 1.2
69 m	1 min	0:59	Loop, set point 1.2
66 m	2 min	1:01	″
63 m	2 min	1:03	″
60 m	2 min	1:06	″
57 m	3 min	1:09	″
54 m	3 min	1:12	″
51 m	3 min	1:16	″
48 m	5 min	1:21	″
45 m	5 min	1:26	″
42 m	5 min	1:32	″
39 m	8 min	1:40	″
36 m	8 min	1:48	″
33 m	8 min	1:57	″
30 m	12 min	2:09	″
27 m	13 min	2:22	″
24 m	17 min	2:40	″
21 m	11 min	2:51	EAN50
18 m	15 min	3:06	″
15 m	22 min	3:29	″
12 m	30 min	3:59	″
9 m	47 min	4:46	″
6 m	195 min	8:02	EAN80 · 35 min of it on EAN50 breaks
0 m	—	8:02	Surface

What the plan assumes

Model. Bühlmann ZHL-16C with gradient factors of 20/80: deep first stops, the way we planned in 2006.
Depth. Planned to the seabed at 93 m, the deepest point on the site, with 55 minutes from leaving the surface to leaving the bottom.
The loop. Tx 10/70 diluent. Set point 0.7 on the way down, 1.2 on the bottom and all the way up to 24 m.
Off the loop at 21 m. The scrubber sets the limit: the axial canister was advised for about four hours, and this is cold water. From 21 m it's open circuit from the stations: EAN50, then EAN80 at 6 m with five minutes on EAN50 in every twenty-five.
Rates. Down at 20 m/min, up at 9 m/min, stops every 3 m, the last at 6 m on the station bar.
Breathing. 15 litres a minute hanging on the stations; 30 on bailout at depth and 20 once on the stops; 1.0 litre a minute of oxygen metabolised on the loop.

$ analyse --o2 --he ~/fills/*

04 / The gas

My part of the job.

Five gases, blended by partial pressure and every cylinder analysed before it was labelled.

Gas	O₂ / He	Used for	Max depth PO₂ 1.4 · 1.6	Min depth PO₂ 0.16	At its depth PO₂ · END · density
Tx 10/70	10/70	Diluent, and bottom bailout	129 · 149 m	6 m	93 m: 1.03 · 21 m · 5.3 g/L
O₂	100/0	The loop's oxygen	—	—	added by the solenoid
Tx 21/35	21/35	Bailout from 57 m	56 · 66 m	surface	57 m: 1.41 · 34 m · 6.1 g/L
EAN50	50/0	Station, 21 m to 9 m, and the breaks at 6 m	17 · 21 m	surface	21 m: 1.56 · 21 m · 4.1 g/L
EAN80	80/0	Station, the long hang at 6 m	7 · 9 m	surface	6 m: 1.29 · 6 m · 2.2 g/L

What the loop is breathing

Where	PO₂	O₂ / He / N₂	END	Density
On the bottom, 93 m	1.20	11.6 / 68.7 / 19.6 %	22 m	5.4 g/L
Deep stops, 57 m	1.20	17.9 / 63.9 / 18.2 %	14 m	4.0 g/L
Last stop on the loop, 24 m	1.20	35.2 / 50.4 / 14.4 %	7 m	2.6 g/L

END counts oxygen as narcotic. Gas density is hindsight: nobody was counting grams per litre in 2006. Current guidance (Anthony and Mitchell, 2016) is to stay below about 5.2 g/L and never go above 6.2; the loop on the bottom comes out at 5.4, and the Tx 21/35 bailout at its switch at 6.1.

The blending sheet

Gas	Fill	Helium to	Oxygen to	Air to
O₂	200 bar	—	200.0	—
Tx 10/70	200 bar	140.0	149.4	200.0
Tx 21/35	200 bar	70.0	88.6	200.0
EAN50	232 bar	—	85.2	232.0
EAN80	232 bar	—	173.3	232.0

From an empty cylinder: helium first, then oxygen, then topped with air, all in bar. These are ideal-gas figures. Helium doesn't behave ideally at those pressures, so in practice it goes in a few bar over, the cylinder settles, and nothing leaves the fill station until it's been analysed for oxygen and helium.

What went in the water

Where	Gas	Cylinder	How many
Megalodon, onboard	O₂	3 L at 200 bar	1 each
Megalodon, diluent	Tx 10/70	3 L at 200 bar	1 each
Sling, bailout	Tx 10/70	11 L at 200 bar	1 each
Sling, bailout	Tx 21/35	11 L at 200 bar	2 each
Station	EAN50	12 L at 232 bar	29 for the team
Station	EAN80	12 L at 232 bar	22 for the team

$ awk '{ sum += $litres } END { print sum }' fills.log

05 / How much

51 station cylinders a dive.

The loop is frugal. The stations aren't: five hours of open circuit for eight divers is a lot of gas, and it all had to be blended between dives.

Station gas

Gas	A diver	The team	Plus half again	Cylinders
EAN50	5,160 L	41,270 L	61,900 L	29 × 12 L at 232 bar
EAN80	3,830 L	30,670 L	46,010 L	22 × 12 L at 232 bar

Litres at the surface, at 15 litres a minute, for 8 divers, with half as much again in reserve; each cylinder counted down to 50 bar. On the loop, by contrast, a diver metabolises about 160 litres of oxygen in 2 h 40 min, against 540 usable in the 3-litre cylinder; the scrubber runs to 67% of its advised four hours.

Bailout

Gas	Needed	Carried	The pair
Tx 10/70	2,990 L	1,980 L	3,960 L
Tx 21/35	6,620 L	3,960 L	7,920 L

The case that sizes it: the loop lost at the end of the bottom time, and a full open-circuit ascent to the stations. Worked through, your own sling cylinders cover that for the first 35 minutes on the bottom. After that it takes your buddy's cylinders as well, and between them a pair covers one diver's bailout up to 62 minutes — 7 more than the plan. It's the reason a buddy pair stays within reach on a dive like this.

$ watch -n 60 cns

06 / Oxygen

Past the NOAA table, and why.

The oxygen clock was built for single exposures of a few hours. Eight hours at useful decompression oxygen levels doesn't fit inside it.

When	Runtime	Depth	CNS	OTU
Leaving the bottom	0:55	93 m	25%	69
Off the loop	2:40	21 m	75%	208
Arriving at 6 m	4:47	6 m	139%	365
Surfacing	8:02	0 m	234%	624

The plan finishes at 234% of the NOAA single-exposure limit; no eight-hour decompression at useful oxygen levels stays inside it. What the plan can control, it does: the loop held at 1.2 rather than 1.3; EAN80 instead of pure oxygen for the long hang at 6 m, which on its own roughly halves the figure; and five minutes on EAN50 in every twenty-five. The pulmonary dose, 624 OTU, stays inside the 850 OTU single-day limit.

$ diff plan.txt plan-b.txt

07 / If it went wrong

Plan B, and plan C.

Same model, same gases, same gradient factors.

Case	Bottom	First stop	Deco	Runtime	CNS · OTU
As planned	55 min	69 m	6 h 57 min	8:02	234% · 624
Five minutes longer on the bottom	60 min	72 m	7 h 34 min	8:44	254% · 676
Loop lost at the end of the bottom time: open-circuit bailout	55 min	72 m	6 h 40 min	7:45	210% · 557

The bailout comes out shorter than the plan because open-circuit Tx 21/35 carries more oxygen through the deep stops than a loop held at 1.2. It costs far more gas, which is the table above.

Not a plan to dive.

Bühlmann ZHL-16C (compartment 1b) with gradient factors of 20/80, in sea water, 10 m to the bar. The engine is bin/build-dive-plan.php on this site; it produces the same stops, minute for minute, as the open-source DecoTengu planner on eight reference profiles. It's here to show how the dives worked. Nobody should dive from it.

RMS Lusitania: the site →

the log: 2006 →

→ diving · dive log 2006 · off duty