Craig Roberts and 2 others · Prevention of Future Deaths report

In July 2013 I commenced an investigation into the deaths of Craig Roberts, James Dunsby and Edward Maher. The investigation concluded at the end of the inquest on 14th July 2015. The conclusion of the inquest was a narrative as per the attached record of inquests.

• Craig Roberts, Edward Maher and James Dunsby were all reserve soldiers who were taking part in a selection process for a specialist unit being held at Brecon Beacons in South Wales in July 2013.
• Reservists have to successfully complete a part time 6 month selection pathway made up of two phases, the aptitude selection (leading up to test week), and then for successful candidates, a continuation phase. Aptitude training started with an induction weekend followed by 8 training weekends culminating in a 2 week training camp at Sennybridge.
• Induction weekend involved a medical, briefings and fitness test.
• The first 2 training weekends comprised of hill preparation inc briefings on map reading, skills, medical matters and kit requirements. Hills preparation weekend concluded with a combat fitness test in uniform, wearing boots and carrying a weight to be completed in a set time.
• The remaining 6 weekends involved a series of day and night marches in the same area as test week. Reserve Units (RU)1 and 2 combine for these activities.
• Only 20 out of 67 candidates from RU 1 got through to test week.
• Most reservists arrived at Camp in Wales on 04/07. The first week of the camp was spent by reservists as follows: 05/07 6 mile run. 06/07 heat injury presentation by 1U. 07/07 a march over 24 km was to be complete within 4hrs 45 mins – due to the weather reservists adjusted the march, reduced the weight to be carried, provided additional water at Pen Y Fan and Windy Gap for drinking and emergency cooling, carried the march out in groups, removed the time limit and provided sweepers. The next 4 days were military skills. 12/07 was a rest day with a final briefing and handover to Signals Regiment who were responsible for test week.
• The Signals Regiment run their own course open to a variety of regiments. The course includes an aptitude phase and test week.
• The Signals Regiment are a regular army unit and they were able to train for a continuous two week period in the run up to test week in the relevant training area undertaking “Lodestone” marches. Distance and weight to carry were increased over time against speed. They were acclimatising to the environment and conditions in the build up to test week. Notably the week before test week saw increasing temperatures from 21.4C on 05/07 up to 26.3 on 12/07 (temp from Sennybridge as this is the best available data).
• For the test march on 13th July 2013 there were 37 reservists made up of reserve units 1 and 2 and 41 regular troops from the Signals Regiment.
• Directing staff (DS) for the march were made up of Signals regiment and RU staff.
• Test week briefing to DS and candidates was undertaken by soldier 1B on 12/07/13.
• Test march covered 26.4km (as the crow flies) through five check points, actual distance estimated as 29-30km depending on the route taken. Candidates had to carry a bergan weighing no less than 49lbs, not including food or water. They were required to carry 3 litres minimum of water. They had to carry a dummy rifle. They were expected to finish the course within 8 hours and 48 minutes.
• Four different routes were allocated to soldiers – red, black, green and orange. Black and red routes started from checkpoint 1 going in opposite directions and green and orange from checkpoint 4 going in opposite directions.
• Water was available at checkpoints (CP) 1, 4 and 5 – CP’s with vehicle access.
• The distances between check points where water was not available were CP4 – 3 – 5 was 12.67km, CP 4 – 2 – 1 was 9.8km.
• Weather forecast for the day was available at the control room at the camp. The met office forecast was a Wet bulb globe temperature (WBGT) of 25 and max speed of 10knots. Many news agencies were reporting that it was forecast to be the hottest day of the year with temperatures predicted to reach 27 degrees. Further WBGT tests were done at 08.00, 12.00 and 16.00 at Sennybridge camp. The result at 12.00 on 13/07/13 was 31.2. There was no WBGT equipment on the march and this result was never asked for nor communicated to the hills.
• All soldiers were issued with a GPS tracker device kept in the top section of their bergans. The location of the trackers could be monitored on computer screens within a control vehicle manned by Signals Regiment staff based at checkpoint 1 – 1A, 1B and 1C. The tracker refreshed every 10 minutes on the control screen. Neither the slow nor the static functions were enabled.
• Tracker devices are equipped with an emergency button which if pressed would activate a signal to the control vehicle.
• Candidates set off at two minute intervals from the checkpoints with Roberts and Maher setting off on the black route and Dunsby on the red route.
• Maher set off at 06.46 Black route 4
• Roberts set off at 06.56 Black route 9
• Dunsby set off at 06.52 Red route 7
• Checkpoint procedure was for candidates to approach and within 10 to 15 yards to drop to one knee with map and compass in hand. Each candidate was then called forward by the directing staff manning the checkpoints and asked for his call sign and route number and where he has just come from and he will then be given the coordinates for the next checkpoint and will show the directing staff where he is going on the map. If the directing staff were happy, the candidate would continue.
• The means to withdraw were – voluntary withdrawal – which meant the soldier could not have a further go at selection, medical withdrawal – the soldier could have another go at test week, training officer withdrawal – the soldier may have another go at test week depending on the reason for withdrawal.
• During the day there were a number of heat related casualties. Soldier 2J medically withdrew at CP4 at 12.14pm with heat illness and Soldier 2P medically withdrew at CP1 at 12.46pm with heat illness, 1W assessed at CP4 for heat illness at 12.22 but was allowed to continue. Soldier 4E medically withdrew at 14.26 with heat illness. 1W was identified as slow to progress at 15.45 and directing staff went to him from CP2. He was assisted down the mountain and found to be suffering from heat illness. He was hospitalised. 1X had his man down alarm triggered at 16.55. He was evacuated by air ambulance to hospital. 2D made it to the finish but later collapsed at camp and had to insist that he was taken to hospital where he was diagnosed with acute kidney injury due to heat illness. 4G was an unidentified heat illness casualty – he finished the course but was unwell afterwards though he did not seek medical assistance.
• The chronology for each soldier is set out in the attached record of inquest.

(1) A new tracker system has been introduced recently. The new system’s slow man/static function does not work. It is therefore still the case that those running the exercise have no means to identify static or slow moving soldiers.
(2) Those in a senior commanding position were unaware that the new tracker system’s slow man/static function did not work until the inquest – this came to light as I asked for a demonstration of the new system which was undertaken on the Malvern’s on Sunday 21 June 2015. I was informed the slow man/static function did not work. I heard no evidence that any steps have been taken to address this problem and no interim measures have been put in place to mitigate the risk.
(3) Witnesses at the inquest confirmed that before this tragedy they were unaware of the main guidance for heat illness namely JSP539– joint service code of practice - climatic illness and injury in the armed forces Version 2:1 November 2012. Some witnesses in a very senior position – AA and SR44 – claimed, respectively, that this guidance was not applicable to this endurance exercise. Others ████████ and ████████ confirmed it was the current guidance and no separate guidance had been issued for exercises with this specialist group. I am concerned that the MOD still do not have a clear plan and guidance for the detection of heat illness in this type of exercise and have failed to instruct commanders of the importance of adhering to JSP539 for this type of activity.
(4) Senior commanders had received no training before this tragedy on JSP539. There was no clear system for disseminating information to different regiments and no means to check those commanding this type of exercise had the requisite knowledge and training.
(5) Senior commanders were unaware that the staff who completed the risk assessment for this exercise and who conducted the exercise had not been trained in the preparation of risk assessments. The risk assessment used simply adopted a risk assessment that had been prepared by the lead regular unit.
(6) Senior commanders were unaware that the staff who conducted this exercise were unaware of climatic guidance in JSP539 and therefore did not understand the implications of the weather forecast and the importance of heat illness and its treatment.
(7) Senior commanders were unaware that the reservist units had a different build up to test week. The reservists had a military skills week the week before test week whereas the signals regiment had build up marches. None of the signals regiment students suffered heat illness.
(8) The general system for reporting heat illness cases is disjointed and results in cases being missed and therefore not reported. Inaccurate data impedes the ability of the MOD to assess the true incidence of heat illness during exercises and to put in place any plan that’s required to mitigate ongoing risks of heat illness.
(9) The tracker system used at the time was known to be unfit for purpose in that the slow man/static function did not work effectively. No commander at any level addressed this deficiency in any directions to staff or further risk assessments.
(10) A previous fatality, Soldier G see LAIT report October 2012, had identified that treatment for casualties should be within the “golden hour”. In addition following Private Poole’s death in 2009 it was identified that the tracker was not fit for purpose and standard operating procedures were issues dated January 2011. None of these recommendations were implemented by those involved in this exercise. I am concerned that lessons had not been learnt from these previous events. There appears to be no clear pathway for communicating this sort of information.
(11) There is no system in place to ensure that WBGT readings obtained at Sennybridge camp are communicated to exercise commanders in the area during the day.
(12) There was no involvement of a doctor experienced in heat illness detection and treatment when devising the medical plan for this exercise; the medical plan was prepared by a junior combat medical technician.
(13) There was no prior liaison with the NHS and Mountain rescue before this exercise about what their involvement might be in the case of any injuries or illness.