Skip to main content

The Manchester Arena Inquiry has now concluded. The closure notice from the Inquiry Chairman is available here.

Volume 2 is divided into two sub-volumes: Volume 2-I and Volume 2-II. Volume 2-I is 695 pages long. Volume 2-I begins with a Preface and then continues with Parts 9 to 16. Volume 2-II is 189 pages long. It contains Parts 17 to 21 and the Appendices. A list of the names of the twenty-two who died is at page vii of Volume 2-I and at page iii of Volume 2-II.
A large format version combining Volume 2-I (ia, ib and ic) and Volume 2-II is also available.
Volume 2-I (standard format)
Volume 2-II (standard format)
Volume 2 (large format)

Exercise Winchester Accord and other exercises

Key findings

  • There was a well‑run programme of multi‑agency exercising in Greater Manchester.
  • There was good participation in multi‑agency exercises by Category 1 responders.
  • Control rooms for the emergency services were not sufficiently involved in relevant multi‑agency exercises.
  • There was a failure to capture lessons learned accurately, or sometimes at all, from multi‑agency exercises.
  • There was a failure to implement change based on what was revealed by multi‑agency exercises.
  • The failure to include North West Fire Control (NWFC) in Exercise Winchester Accord was not sensible. It was a missed opportunity to allow NWFC to get important experience of mobilising resources to a Major Incident.
  • Exercise Winchester Accord was too large to be run as a regional, Tier Three exercise. The national interest in it was not matched by the required planning and support to capture the learning from such a large and complex exercise.
  • The scale and scope of Exercise Winchester Accord’s objectives did not allow for issues to be identified and lessons learned in an effective way. There was a tension between the different objectives for different organisations.
  • Greater Manchester Police’s decision to look at how the Force Duty Officer (FDO) operated in a set‑up that was different from the existing one during Exercise Winchester Accord risked taking attention away from the well‑known concerns about the FDO role.
  • The local objectives set during Exercise Winchester Accord for the FDO were wide enough to look beyond the proposed move and to test the well‑known issues with how the role worked.
  • The draft action cards were not tested during Exercise Winchester Accord.
  • This was a significant missed opportunity to test and improve known weaknesses in the role of the FDO and the capabilities of the Operational Communications Branch during a Marauding Terrorist Firearms Attack.
  • The debrief process on Exercise Winchester Accord was flawed.
  • Exercise Winchester Accord represented a significant missed opportunity to prepare an adequate and robust response to a Marauding Terrorist Firearms Attack within Greater Manchester.


There were many different types of multi‑agency exercise organised in Greater Manchester before May 2017. It is not necessary to include reference to them all in this Report. In the next section, I will analyse one in particular, Exercise Winchester Accord, in detail.

Before I do, it is necessary to say something about exercising more generally.

Defining an exercise

An exercise is a simulation of an emergency situation. It helps to check and validate plans. It allows people to practise carrying out their roles and to test well‑established procedures.980

The two types of exercise relevant to this Inquiry were: tabletop and live exercises.981

A tabletop exercise is based on a realistic scenario and timeline. The timeline may be in real time or it may be speeded up. Usually, tabletop exercises are run in a single room. To simulate the divisions between responders who need to communicate and be co‑ordinated, they can be run in linked rooms. The players are expected to know the plan, and they are invited to test how the plan works as the scenario unfolds.982

An example of a multi‑agency tabletop exercise was Exercise Sherman. I considered Exercise Sherman in Volume 1. It tested a multi‑agency response to a Marauding Terrorist Firearms Attack. Inspector Roby described how the format of Exercise Sherman was to distribute attendees between pre‑assigned tables.983 Attendees were from local businesses and emergency services. Everyone was encouraged to participate with those on their table and on other tables. There were discussion points, and observations were invited. There was a plenary session with an open invitation to make further comments and raise questions.984

At the conclusion of Exercise Sherman, attendees were asked to complete a feedback form. All feedback was captured on an Exercise Recommendation Tracker and discussed at the GMRF Resilience Development Group.985 This was an example of a well‑organised, inclusive exercise.

A live exercise is a rehearsal for implementing a plan. Such exercises are particularly useful for testing logistics, communications and capabilities. Guidance on the GOV.UK website indicated: “Live exercises are expensive to set up on the day and demand the most extensive preparation.”986

An example of a live exercise was Exercise Lionheart. This was a series of night‑time exercises at the Arndale Centre in Manchester in April and May 2015.987 Two of the exercises were multi‑agency. The multi‑agency exercise objective was to test the Operation Plato response to a Marauding Terrorist Firearms Attack and the application of JOPs 2 by commanders from GMP, GMFRS and NWAS.988

In Greater Manchester, there were three broad methods of organising the different types of multi‑agency exercise which were tested. First, those initiated by GMRF, such as exercising of flood plans. Second, national exercising where GMRF and partner agencies played a role, such as Exercise Winchester Accord. Third, exercises led by an individual agency which other agencies participated in, such as an exercise on responding to a chemical spill organised by GMFRS.989

Multi-agency exercises: the positives

I was assisted on the issue of multi‑agency exercising by evidence from all the Emergency Response Experts, from GMRF and from corporate witnesses on behalf of the emergency services.

All of the Emergency Response Experts agreed that the multi‑agency exercise regime co‑ordinated by GMRF was well structured and the opportunities available to each service were very good.990

There was good participation in multi‑agency exercises by emergency services in Greater Manchester. NWAS held subject‑specific multi‑agency exercises for a Marauding Terrorist Firearms Attack scenario every 6 to 12 months.991 GMP participated in at least a hundred exercises in the two years before the Attack. A number of these were multi‑agency exercises.992 As the Ambulance Service Experts observed: “[T]his demonstrates a high level of commitment to the training and exercising obligations placed on a Category 1 responder.” 993

Exercising is very expensive. A live exercise will likely involve hundreds of participants and a substantial investment of time to plan and conduct the exercise. The programme of multi‑agency exercising in Greater Manchester was maintained despite significant budget reductions, particularly for GMP.

As I have already said, from 2010/11 to 2017/18, GMP income fell by over 23 per cent, and the number of its police officers fell by 25 per cent.994

As all the Emergency Response Experts noted, GMP, GMFRS and NWAS recognised the real importance of exercising. They dedicated an appropriate level of resource, time and commitment to it.995

Multi-agency exercising: the problems

There was a failure to include control rooms in multi‑agency exercises sufficiently, or sometimes at all. This was exemplified by the failure to include NWFC in Exercise Winchester Accord.996 During Exercise Lionheart, NWFC was simply informed of the exercise dates and the possibility of gunfire and explosions to avoid the risk of appliances being mobilised.997

BTP should have been included in the programme for Greater Manchester multi‑agency exercising but was not. The Policing Experts concluded: “Control Room structures, arrangements and the training of staff in a response to serious emergency may have been compromised.” 998 This is an assessment with which I agree.

Although considerable effort was made to maintain a schedule of multi‑agency exercises, there was a failure to derive and embed learning adequately from some important exercises. There was a lack of scrutiny of exercise objectives against performance. This was a failure by GMRF to ensure that there was a robust debrief process in place.999

There was no comprehensive system for maintaining records of exercises or details of who attended. For example, DCC Pilling explained that the exercise records held by GMP were “fragmented”. He stated that it was “difficult to say with complete accuracy exactly how many exercises have been delivered to officers over a period of time”.1000

In the GMFRS feedback from Exercise Lionheart, Group Manager Levy identified that delay in declaring an Operation Plato warm zone resulted in the delayed deployment of Marauding Terrorist Firearms Attack responders.1001 The College of Policing debrief for the same exercise did not identify this as an area for improvement. Instead, under a heading of “[P]erceptions of What Went Well”, the feedback commented on “introducing a ‘WARM ZONE’ as early as possible”.1002 It was said that this allowed HART and GMFRS to enter the training area early and was good for training the evacuation of casualties by air, or “casevac training”.1003 This was a failure to understand and capture an important area of feedback. The issue of zoning arose again in Exercise Winchester Accord 12 months later and during the emergency response to the Attack a year after that.1004

The Policing Experts commented, as there was no comprehensive system for monitoring exercises, it was difficult to understand how organisations could be sure that lessons were learned.1005 Sometimes lessons were not captured at all. Sometimes lessons were simply not recorded accurately.

This meant that, despite the commitment to planning and participating in multi‑agency exercises, critical aspects of the learning that should have been identified were not. This was not an isolated error but occurred repeatedly.

In the GMRF multi‑agency debrief following the Attack, a number of problems were identified which had arisen during earlier multi‑agency exercises.1006 First, there was the lack of communication surrounding the declaration of Operation Plato. A similar issue was identified during Exercise Winchester Accord. Second, there was early confusion concerning the establishing of RVPs and the FCP. A similar concern also arose on Exercise Winchester Accord. Third, there was the need for an Airwave talk group to enable the sharing of risk‑critical information in real time, a problem that was flagged on Exercise Hawk River.1007

The Ambulance Service Experts explained that the importance of joint understanding of risk and shared situational awareness among co‑located police, ambulance, and fire and rescue commanders was a theme of multi‑agency exercises in 2015 and 2016. They gave examples of Exercise Dawn Vigil in July 2015, a Counter Terrorism Policing exercise in October 2015 and Exercise Lawman 2 in March 2016.1008 Despite the learning that there were risks in the approach to joint situational awareness, these were issues that arose again in the emergency response to the Attack.

Exercises uncover problems and identify better ways of working. That is their purpose. It is important not to apply what is now known happened on 22nd May 2017 to the approach to multi‑agency exercising from before the Attack. There was a good programme of multi‑agency exercising, but there were nonetheless problems with it. Most significant was the inability to identify, record and respond to lessons learned. In the future, a system must be put in place to address this. A candid approach to learning is vital to ensure agencies can work together effectively.

With those comments in mind, I turn now to Exercise Winchester Accord. A number of the people who participated in Exercise Winchester Accord were involved in the emergency response on the night of the Attack. Some were promoted between the two events. Some have been promoted since. As it was an exercise, rank during Exercise Winchester Accord is less important to my conclusions. To avoid confusion, I shall refer to individuals by their rank as at 22nd May 2017, not their rank at the date of the exercise.

Aims and objectives of Exercise Winchester Accord

On 2nd November 2015, GMP received a request to host Exercise Winchester Accord in Greater Manchester.1009 I heard a significant amount of evidence about Exercise Winchester Accord, but my investigation into it was not exhaustive or in the same level of detail as was the case for the events on 22nd May 2017.

To many, it foreshadowed critical failures in the emergency response to the Attack. In particular, the overburdening of the FDO, the failure to communicate a declaration of Operation Plato to the fire and rescue and ambulance services, and the failure to establish a joint FCP.1010 This view was not universally shared, particularly by GMP. GMP did not consider that the exercise showed a catastrophic failure of the FDO or delays at the FCP. GMP cautioned against drawing comparisons between the exercise and the events on 22nd May 2017.1011

Exercise Winchester Accord took place over three days from 9th to 11th May 2016. It was a “live-play” exercise with over 1,000 “players” and 160 “casualties”.1012 It was conducted in three phases. My focus has been on the first phase: the Marauding Terrorist Firearms Attack scenario at the Trafford Centre in Manchester.1013 This commenced at 00:00 and went on into the early morning of 10th May 2016.1014

The primary aim of Exercise Winchester Accord was the relicensing of military assets for domestic purposes.1015 Additionally, GMP and other agencies were invited to set their own objectives for the exercise.1016 Inspector Roby was the exercise co‑ordinator. She led the discussions aimed at setting the local objectives.1017 She described how those objectives were “bolted on” to test particular issues.1018 She explained that she was nearing the end of writing the plan for the Strategic Co‑ordination Centre, which became known as the Force Command Module, and the exercise was a prime opportunity to test it.1019

About 70 different local objectives were set.1020

NWAS objectives included demonstrating the effective evacuation of patients from the Operation Plato warm zone to a Casualty Clearing Station, testing communication links between NWAS commanders internally and with the North West Counter Terrorist Unit (NWCTU) operations room, and demonstrating the ability to provide appropriate clinical care to ballistic injuries sustained in a Marauding Terrorist Firearms Attack.1021

GMFRS objectives included testing the integration of the NILO with the Counter Terrorism Commander, testing the use of joint dynamic risk assessment in line with JESIP, testing the GMFRS management at a Marauding Terrorist Firearms Attack in line with JOPs and the wider incident support GMFRS could offer.1022

GMP set over 50 force‑specific objectives.1023 These were wide ranging and included local objectives for the GMP Operational Communications Branch where the FDO worked, the Operational Planning Unit, the Press Office, Scene Management and partner agencies in a Strategic Co‑ordination Centre.1024 More details about the FDO objectives will be set out at paragraphs 12.779 to 12.792. Separate objectives were also set for NWCTU and regional armed policing.1025

Role of NWFC during Exercise Winchester Accord

Despite the breadth and ambition of local objectives, NWFC was not invited to participate in the exercise.1026 It played no role in mobilising any GMFRS resources. Rather, the NILO was given the role to mobilise GMFRS to the exercise.1027 NWFC was simply made aware that the exercise was happening so that it did not impact on its own management of fire resources while the exercise was under way.1028

In an email dated 25th March 2016, Group Manager Levy explained to NWFC Operations Manager Janine Carden that he had enquired about NWFC involvement in the exercise, “both for operational / logistic mobilising, and also as a training / assurance opportunity for you”.1029 Group Manager Levy said that he recognised that there was a clear need for NWFC involvement in mobilising resources and engagement in multi‑agency communications.1030

There was a suggestion that staffing levels in the NWFC control room played a part in the reason NWFC were not involved in Exercise Winchester Accord. This was incorrect.1031 Station Manager Gaskell explained that the view was that Exercise Winchester Accord did not “lend itself to a dynamic mobilisation exercise”.1032 He stated that there was no benefit to NWFC being involved as the starting point of the exercise was from the pre‑arranged RVP.1033

During the Attack, just 12 months later, the breakdown in communication between the duty NILO and NWFC was a significant feature of the failures to deploy GMFRS personnel to the Arena. In that context, the failure to involve NWFC in a large, multi‑agency exercise such as Exercise Winchester Accord was not sensible. It was a missed opportunity to allow NWFC, a relatively new organisation, to get important experience of mobilising resources to a Major Incident.

Planning of Exercise Winchester Accord

Despite the failure to include NWFC, Exercise Winchester Accord was an ambitious exercise. Inspector Roby said that, in planning it: “We were desperate to exercise a lot of structures that we had not had a chance to.”1034 An NWAS planning document for the exercise described it as “one of the largest staged in the UK”.1035 Yet, despite its size, Exercise Winchester Accord was organised as a Tier Three exercise.1036 This meant that the exercise was organised, run and debriefed at a regional level, not nationally.1037

The Policing Experts observed that a Tier Three exercise did not have the “same support mechanism” compared with a Tier One national exercise.1038 A Tier One exercise would receive central government support, and there would be a rigorous focus on the evaluation of each exercise objective.1039 In contrast, in a Tier Three exercise the debrief process would be more open and generic, without evaluating specific objectives.1040

The Policing Experts believed that Exercise Winchester Accord was probably too large to be managed as a Tier Three exercise.1041 This is a view with which I agree. The NWAS planning document noted: “[T]here is national interest in the outcomes.” 1042 That national interest was not matched by the required planning and support to capture the learning from such a large and complex exercise.1043

Exercise Winchester Accord had a mix of national and local objectives.1044 Inspector Roby acknowledged that planning and policy‑making were “hit very badly” by budget cuts from 2011.1045 She said it impacted on the ability of her planning team to do their jobs.1046 CI Booth also observed that there was a reduction in staff numbers at the Operational Communications Branch because of budget cuts. He said it was a “very demanding time” and staff were under pressure.1047

It is understandable therefore that, where there was an opportunity to plan a large, live exercise, there was a temptation to include many different objectives. As Inspector Roby acknowledged, she “threw everything at it”.1048 It was the responsibility of GMRF and the agencies involved in planning multi‑agency exercises to plan exercises in an effective and coherent way.1049 An organisation needed to have overall responsibility for the conduct and the content of the exercise. Even now, there is not agreement as to who this was for Exercise Winchester Accord. As was shown by the confusion in the evidence about who was responsible for organising, participating in and reviewing Exercise Winchester Accord, the scale of the objectives did not allow for issues to be identified and lessons to be learned in an effective way.1050

This was exemplified by the evidence of the GMRF Chair, Deputy Chief Fire Officer Argyle. On the first occasion he gave evidence, he referred to Exercise Winchester Accord as two exercises. He said that GMRF was involved in a separate exercise to test a Strategic Co‑ordination Centre.1051 This was explored again with Deputy Chief Fire Officer Argyle when he gave evidence for the second time. He said he thought that Exercise Winchester Accord was an “exercise running two separate ways”.1052 Deputy Chief Fire Officer Argyle said that the exercise provided the opportunity to test the plan of GMRF members in setting up a Strategic Co‑ordination Centre at GMP HQ and how a Strategic Co‑ordinating Group would run.1053

Exercise Winchester Accord was one exercise, albeit large and with involvement from many different organisations, each testing different and sometimes overlapping issues. The fact that the Chair of the local resilience forum was not able to say with certainty whether it was one exercise or two and was focused on a particular aspect of it, suggests that there were problems in how large‑ scale, multi‑agency exercises were managed. In the future, more resources and robust processes should be put in place where large, regional exercises are planned with multiple objectives.

Force Duty Officer objectives during Exercise Winchester Accord

Ten local objectives were set to test the role of the FDO.1054 These included: identifying information leading to the correct declaration of Operation Plato; ensuring Operation Plato protocols were followed; identifying sufficient command and control structures to deal with an ongoing incident; notification protocols within GMP and to other agencies; and examining the structures for the proposed relocation of the FDO to GMP HQ.1055 Some other GMP local objectives also appeared to touch on the role of the FDO, such as an objective for examining the immediate command and control to establish who was informed of an incident and who had operational command.1056

Inspector Roby explained that a purpose of the FDO objectives was to replicate what would happen in real life.1057 She spoke to relevant people, including an FDO, CI Booth, who was working at a senior level in the Operational Communications Branch, and Laura Lewis, who was the control room manager.1058 Inspector Roby said she asked them, “what was going to cause problems if we had this type of incident in reality”.1059 The answers fed into the type of objectives that were set.1060 They contributed to the sequencing of the exercise.

The Marauding Terrorist Firearms Attack role play began at 00:00. A minute later, calls were to be placed from role‑playing shoppers, residents and others to play out the scenario of an active terror attack. Calls to the FDO telephone line were scheduled to continue for a further 19 minutes.1061 The calls were scripted and included press queries, and calls from GMP officers, other police and emergency services, members of the public and others.1062 

The sequence of events for the exercise noted that, by 00:03, three minutes after the simulation started, callers “[m]ay struggle to contact FDO due to number of phone calls coming in. Need to keep trying.”1063 The FDO was expected to declare Operation Plato at 00:09, while the calls were being received.1064 Inspector Roby said this was “to replicate the sheer volume of calls that would be coming in”.1065 She emphasised that the FDO’s telephone line was like a switchboard system with flashing lights.1066 It was expected that other staff would be in a position to answer calls.1067

Inspector Roby also said that the set‑up of the FDO during Exercise Winchester Accord and during a real Major Incident at the time were not comparable.1068 She explained that the exercise was not meant to be a real test of the FDO’s capacity in the way the role was performed at the time.1069 Inspector Roby said that it was a test of proposed capacity in the event that the FDO moved to GMP HQ.1070

Inspector Roby said she did not think that proposal would work.1071 She believed that, if the FDO were moved out of the Operational Control Room, it would vastly reduce their ability to delegate.1072 She said that, if the FDO were working remotely, they would not get a feel for the room; they would have to be told that something was happening.1073

CI Booth, who was involved in planning the exercise, similarly said: “Winchester Accord, from an FDO perspective, had been about testing the FDO in a new location, FHQ [Force Headquarters – GMP HQ], with reduced support.”1074 It was proposed that there would be a small team of radio operators with the FDO working from an area within GMP HQ called ‘the Force Hub’.1075 CI Booth was concerned that, by reducing the number of staff, the “criticality around being overwhelmed was more likely to bear fruition [sic]”.1076 As a result, a local objective was added to Exercise Winchester Accord to test the proposed Force Hub.1077

Although not directly involved in Exercise Winchester Accord, DCC Pilling explained that it was not a test of the FDO working in the Operational Control Room with the level of support which would have been available on the night of the Attack. Rather, the exercise placed the FDO in an unfamiliar environment at GMP HQ, without the support of the wider Operational Communications Branch.1078

During Exercise Winchester Accord, the FDO was supported by nine people. Four were radio operators and four were Operational Communications Branch staff.1079 These details can be seen in Figure 35 at the two desks in the lower half of the diagram. During the Attack, the FDO was supported at the Operational Control Room by 34 people.1080 This can be seen in Figure 36. DCC Pilling said that these differences in the support for the FDO, “inevitably impacted on the lessons that were drawn from the exercise about the FDO: they primarily concerned the proposed moved of the FDO to FHQ [Force Headquarters – GMP HQ]”.1081

Figure 35: Force Duty Officer team on Exercise Winchester Accord1082

Figure 36: GMP Operational Control Room as at 22nd May 20171083

Inspector Roby and CI Booth were overly focused on testing a scenario they knew would fail. Their firm views that the move of the FDO to the Force Hub in GMP HQ would not work confused the wider opportunity that the exercise offered to test the role of the FDO and their capacity to deal with a Marauding Terrorist Firearms Attack. The local objectives set for the FDO went beyond simply testing the proposed relocation from the Operational Control Room to GMP HQ. The objectives tested the role of the FDO in a number of different ways, not least around the handling of a declaration of Operation Plato.

Irrespective of whether the core local objective for GMP was to test the role of the FDO at a different location, the recognition of the need to test the role of the FDO and the support around them was logical. It showed that there was an understanding at GMP about the likely pressure that the FDO would be placed under during a terrorist attack, wherever he or she was based.

The decision to look at how the FDO operated in a set‑up that was different from the existing one risked taking attention away from the well‑known concerns about the FDO role.1084 To understand that, it is important to look at Exercise Winchester Accord in its wider context.

GMP accepted that it was “well known”1085 that the FDO would be under pressure during a Marauding Terrorist Firearms Attack.1086 On behalf of Counter Terrorism Policing Headquarters, CI Thomas said that it was “a well-understood fact that the FDO was a potential single point of failure”.1087 This was known before and after Exercise Winchester Accord.

The Policing Experts observed that Exercise Winchester Accord should therefore not have masked what was already known about the vulnerabilities of the FDO.1088 I agree with this view. The local objectives set for the FDO were still wide enough to look beyond the proposed move to the Force Hub in GMP HQ and to test the well‑known issues with how the role worked. It is regrettable that the exercise failed to do so.

Background to the Force Duty Officer role

The phrase ‘catastrophic failure’ was used during the Inquiry’s hearings to reference problems with the FDO during Exercise Winchester Accord. It is an evocative phrase. It was used in questioning witnesses and in closing statements. No witnesses volunteered this phrase by reference to the FDO in Exercise Winchester Accord. Some witnesses were asked to agree whether there was such a failure of the FDO during that exercise.1089 The question of whether or not there was a failure of the FDO and, if so, whether that failure was catastrophic must be approached with care.

In order to understand the role of the FDO for Exercise Winchester Accord, it is necessary to set out first some background to that role.

Between February 2016 and September 2018, CI Booth worked in GMP’s Operational Communications Branch. In that role, he had responsibility for the FDO. This was a role he had performed in the past.1090 He said that it was “undoubtedly” an important skill for an FDO to have the ability to communicate with other emergency services.1091 A key element of the FDO role in a Major Incident involving a firearms deployment was to act as the Initial Tactical Firearms Commander. This meant that the FDO would decide whether the deployment of firearms officers was necessary and, alongside a Tactical Advisor, decide how best to deploy firearms officers to deal with the incident.1092

In February 2016, three months before Exercise Winchester Accord, CI Booth contributed to a National Resilience Capability Assessment. The purpose of this assessment was to examine the capability and capacity to respond to a Marauding Terrorist Firearms Attack.1093 On the assessment questionnaire, it was noted that control room staff had “no specific training around the identification of MTFA [Marauding Terrorist Firearms Attack] attacks” but had received situational awareness inputs around the current threat level.1094 For the FDO, it was said that they had “received limited training … around MTFA scenarios”.1095 CI Booth agreed that this was less than ideal. Prior to the Attack, GMP had done nothing to remedy the situation.1096

In February 2016, at the time of the National Resilience Capability Assessment, the Operational Communications Branch did not have action cards in place to deal with a Marauding Terrorist Firearms Attack.1097 CI Booth was given the task to produce action cards for the Operational Communications Branch that could be used by staff during a Marauding Terrorist Firearms Attack and in other Major Incidents.1098 A rough first draft of the action cards was produced by 12th April 2016, just under four weeks before Exercise Winchester Accord.1099 A further draft was circulated by email to the senior leadership team of the Operational Communications Branch on 26th April 2016, just under two weeks before Exercise Winchester Accord.1100 Feedback on the action cards was invited. CI Booth could not recall that any was received.1101

CI Booth said that, at the stage the action cards were being produced, he knew from his own experience that there was a “distinct possibility” of the FDO not being able to cope during a Marauding Terrorist Firearms Attack.1102 He felt that there was more being asked of the FDO in the initial co‑ordination with firearms officers.1103 The action cards would mitigate some of that risk in the set‑up of communications within the Operational Communications Branch, but there was still “potentially a problem”.1104 It was intended that the action cards would be circulated to communications staff who were working on Exercise Winchester Accord and refined as a result.1105

On 3rd May 2016, the week before Exercise Winchester Accord, CI Booth circulated the action cards to Mark Gallagher, who was responsible for resource management in the Operational Communications Branch. The email asked Mark Gallagher to circulate the action cards to staff working on Exercise Winchester Accord and further requested that feedback be provided following the exercise. On the same day, CI Booth also provided the action cards via email to the Silver Control Room Manager assisting on the exercise. The email again explained that the action cards should be tested during the exercise and thereafter feedback should be provided.1106

There was no evidence that the action cards were tested during Exercise Winchester Accord, and debriefs from Operational Communications Branch staff suggested they were not used.1107 This was a significant missed opportunity to use Exercise Winchester Accord to test and improve known weaknesses in the role of the FDO and the capabilities of the Operational Communications Branch during a Marauding Terrorist Firearms Attack. The focus of GMP staff was to prove the adequacy or not of the proposed move to the Force Hub in GMP HQ rather than to test proposed improvements and better ways of working for the FDO.

Multi-agency response during Exercise Winchester Accord

What happened and when during the critical aspects of the Marauding Terrorist Firearms Attack phase of Exercise Winchester Accord is a matter of contention. During the Inquiry there was particular disagreement about how the role of the FDO functioned and whether there was a delay in deploying NWAS and GMFRS into the Operation Plato warm zone to treat casualties.

The trigger for the multi‑agency response was intended to be the shared declaration by the FDO of Operation Plato. This would be the signal for NWAS and GMFRS to move forward to the FCP.1108 The individual RVPs for the agencies were agreed in advance.1109 The FCP was located in a car park adjacent to the Orient entrance to the Trafford Centre.1110

Superintendent Graeme Openshaw was the Ground Assigned Tactical Firearms Commander for Exercise Winchester Accord.1111 The FDO was Inspector Marcus Williams. He was a very experienced firearms officer but relatively new to the FDO role.1112

The exercise sequence of events indicated that, at 01:00, the Ground Assigned Tactical Firearms Commander would “[t]ake control of staff” at the Trafford Centre and “attend scene, direct firearms assets accordingly”.1113 The sequence of events gave no further detail of plans at the Trafford Centre except for one further entry at 01:40 about checking cordons.1114 It is uncontentious that events on the ground at the FCP did not happen according to the exercise plan. It is not agreed why that occurred.

GMFRS’s view

Station Manager Lawlor and Station Manager Gaskell were involved in planning the exercise. Station Manager Lawlor was the exercise co‑ordinator for GMFRS. Station Manager Gaskell explained that it was a very well‑planned exercise.1115 The planning started in late 2015.1116 Until the evidence was presented to the Inquiry, Station Manager Gaskell said he was unaware that one of the local objectives of Exercise Winchester Accord was to investigate the relocation of the FDO.1117 It was something he felt he ought to have known about, particularly if it was going to have a significant effect on the exercise.1118 This is a sensible observation. It is important that all partners in a multi‑agency exercise are aware of all the objectives.

GMFRS and NWAS were told in advance the location of the RVP and the FCP.1119 In my view, it would have been helpful if Exercise Winchester Accord had required the agreement on an FCP during the exercise rather than having a pre‑determined one.

Station Manager Lawlor said that “the exercise did not run smoothly”.1120 He said this was because neither the FDO nor the Ground Assigned Tactical Firearms Commander were in contact with other agencies.1121 The FDO was provided with contact numbers for the GMFRS NILOs.1122 Station Manager Gaskell explained that this was important as the fire and rescue service do not self‑deploy. GMFRS personnel are taught to “maintain discipline and to wait for instructions from other agencies”.1123

Station Manager Gaskell said that the exercise began at 00:00.1124 He stated that the initial 30 minutes were purely for the police to organise their response to the simulated attack. His expectation was that at 00:30, there should have been a declaration of Operation Plato that came into GMFRS and NWAS so they could progress from the RVP to the FCP to carry out the “function that we trained for excessively”.1125 Station Manager Lawlor said that GMFRS was expecting to be informed by the FDO about events unfolding at the Trafford Centre and to give them the location of the pre‑defined FCP.1126 He accepted that GMFRS was not given a timeframe for specific actions by the exercise co‑ordinator.1127

Station Manager Gaskell explained that, on the night of the exercise, specialist GMFRS resources were put on standby. GMFRS officers attended GMP silver command.1128 Stretford Fire Station was used as the GMFRS RVP and muster point.1129 His expectation was that the police would enter the Trafford Centre, identify a terrorist threat, declare Operation Plato and allocate zones. He thought this would then be shared by the FDO with GMFRS and NWAS who would mobilise from their RVP to the FCP.1130

Station Manager Gaskell stated: “In fact we [GMFRS] were not notified of the declaration of Operation Plato at all.”1131 His evidence was not entirely consistent with the view of Station Manager Lawlor about the order of deployments to the RVP and FCP. Even so, there was an expectation by both of them that three commanders for GMP, GMFRS and NWAS would come together at the FCP and that somebody from the police, ideally the Ground Assigned Tactical Firearms Commander, would meet them to help assess risk so they could move forward to carry out rescues.1132

Station Manager Lawlor explained that there was an agreement between the GMFRS and NWAS exercise players to “run with it” when they were not informed by the FDO about the need to move to the FCP.1133 He said that it would have been a “false response”1134 to move to the FCP without contact from the FDO. However, there came a point when it was recognised that the FDO was not going to inform them about the FCP, and a decision was made to move forward to it.1135

At that point in the exercise, Station Manager Lawlor recalled that either he, or someone from NWAS, contacted the GMP exercise planner to confirm the decision to move forward.1136 Station Manager Lawlor said that this meant there were “very lengthy delays” in deploying resources.1137 He explained that, as they were part of a national exercise of significant importance, they did not want to be the cause of that delay.1138

Station Manager Gaskell stated that GMFRS personnel spoke to NWAS personnel while GMFRS personnel were still at their muster point. He stated it was discovered that NWAS had not had any contact from the FDO either. He went on to state that attempts to reach the Ground Assigned Tactical Firearms Commander, who was believed to be at the Trafford Centre at that time, were also unsuccessful. Station Manager Gaskell’s recollection was that it was at this point that both NWAS and GMFRS decided to move forward.1139

Station Manager Lawlor stated that once NWAS and GMFRS personnel moved forward to the FCP, it was anticipated that they would be met there by the GMP Ground Assigned Tactical Firearms Commander to undertake a joint assessment of risk and define the zonings and limit of exploitation.1140 However, he stated that they were not met by anyone from GMP to carry out the “over the bonnet” co‑location and co‑ordination so they could mobilise and commit resources to the zones.1141

Station Manager Gaskell said that the fact that GMFRS and NWAS did not deploy for “two hours meant that some of the police functions on that exercise had been completed”.1142 He recalled that as GMFRS and NWAS personnel were being deployed, some police officers were walking back from the exercise. It was, said Station Manager Gaskell, “an opportunity lost” to observe fire and rescue and ambulance services’ capability under the governance of the police.1143

Station Manager Lawlor stated that at the FCP, it was not possible to make contact with the Tactical Firearms Commander. This contact at the FCP was, according to Station Manager Lawlor, considered “vital” in order to undertake a joint assessment of risk and identify zones and the limit of exploitation.1144 His recollection was that at the request of GMFRS and NWAS, the Tactical Firearms Commander was directed by an exercise co‑ordinator to make contact with other agencies. There was, it was stated, an overall delay of one and a half hours in NWAS and GMFRS deploying into the Trafford Centre.1145

Station Manager Lawlor said that a Police Inspector at the inner cordon would not allow GMFRS and NWAS to enter.1146 However, Station Manager Gaskell said: “[O]nce the actual commanders got together at the FCP, then the exercise actually went very well.”1147

Station Manager Gaskell said that GMFRS was left “disgruntled” by the experience of Exercise Winchester Accord.1148 He said that GMFRS felt overlooked and that the role of the fire and rescue and ambulance services in a Marauding Terrorist Firearms Attack was not appreciated. Station Manager Gaskell felt that the exercise did not deliver, as GMFRS was not used in a timely fashion or in accordance with national guidance.1149

Group Manager Levy was also present on Exercise Winchester Accord. He acted as an observer to support NILOs in development and to advise on the application of JESIP.1150 He was present with the resources that were not being called forward. At the time, he did not observe and was not aware that the FDO had become overwhelmed. His recollection was more with regard to the location of Incident Commanders and their capability to bring the ambulance and fire and rescue teams to the scene.1151 Group Manager Levy said he was unaware of any problem with the FDO1152 until March 2018, when the independent review of the preparedness for and emergency response to the Attack by Lord Kerslake was published.1153

Group Manager Carlos Meakin was one of the development NILOs who observed Exercise Winchester Accord.1154 He attended the initial briefing for GMFRS personnel at the RVP and then went to the Force Command Module at GMP HQ to observe silver command.1155 Group Manager Meakin said that there was a “substantial delay” of around an hour in deploying GMFRS and NWAS responders into the Trafford Centre.1156 He believed that this was due to the absence of the Tactical Firearms Commander at the FCP, who placed himself in the control room at the Trafford Centre “cutting himself off from face-to-face communications”.1157 This was evidence that Superintendent Openshaw disputed.1158

Group Manager Meakin described a feeling of frustration as a result of Exercise Winchester Accord. He was watching remotely but understood that the communication with the Tactical Firearms Commander was so delayed that it severely impacted the deployment of resources into the exercise.1159 He said that could potentially have resulted in the loss of life in a real‑world scenario.1160 A lot of preparation had gone into the exercise, but it did not play out as expected because of the lack of communication.1161

Area Manager Paul Etches was embedded on the exercise in the Silver Control Room at GMP HQ. His role was to act as a liaison for information‑sharing in a multi‑agency partnership.1162 Area Manager Etches initially said his perception was that there was a “lengthy delay” in asking the fire and rescue and ambulance responders to move forward.1163 He said that GMFRS did not have situational awareness about what activity was due to take place, and when, in order to raise the issue.1164

With the benefit of hindsight, Area Manager Etches said that, rather than “delay”, he thought a better description was that the police moved forward to carry out their primary objective, but GMFRS and NWAS services did not get an opportunity to move behind them to carry out the recovery procedures.1165 He did not think that the communication had gone well between GMP and GMFRS. The problem was a gap in the communications on scene, but there was also an opportunity in the Silver Control Room to have better communications.1166

Group Manager Fletcher was an observer on Exercise Winchester Accord. He was based in the management suite at the Trafford Centre.1167 He said he was “surprised” by the length of time it took to deploy GMFRS and NWAS personnel. He was only able to watch the response on CCTV without any sound. He later found out that a Police Inspector would not permit GMFRS and NWAS crews into the cordon.1168

Group Manager Fletcher agreed that there was a significant JESIP failure. It appeared to show a lack of understanding about GMFRS capabilities.1169 The JESIP lead at Merseyside told him that JESIP did not work on the night.1170 Group Manager Fletcher said that it was fed back to him that the delay at the FCP was a failure of the FDO to call the emergency services forward. He acknowledged that this was not something he saw or heard in the management suite.1171

A key learning point for GMFRS from Exercise Winchester Accord was the need for multi‑agency commander training. Station Manager Lawlor explained that there was no one in GMP trained to act as a liaison with GMFRS and NWAS.1172 Station Manager Gaskell said that the focus of this further training was to educate police commanders on the role of specialist responders from other agencies through the JOPs commander course that was introduced in January or February 2017.1173 Station Manager Lawlor said the training was also targeted at GMFRS and NWAS NILOs, the Tactical Firearms Commander, and commanders and Tactical Advisors from all three agencies.1174

NWAS’s view

The NWAS planning document for the exercise set out an anticipated timetable of events. On day one, it was planned that the FDO would contact the NWAS Exercise Co‑ordinator, Paul Bailey, to declare Operation Plato at approximately 00:05.1175 HART crews, the Ambulance Intervention Team Commander and the Operational Commander were to mobilise to the Trafford Centre at 00:25. At 01:00, NWAS senior commanders were to support the Tactical Co‑ordinating Group and Strategic Co‑ordinating Group.1176 These timings broadly corresponded with those set out by GMFRS.

NWAS’s view on the Marauding Terrorist Firearms Attack phase of Exercise Winchester Accord was summarised in its exercise report issued on 3rd August 2016.1177 NWAS’s exercise report noted that there was on‑scene co‑location between GMFRS and NWAS commanders, but that the “lack of police presence did not allow for a JDR [joint dynamic risk assessment] to take place”.1178 These views were repeated in an undated document entitled ‘Winchester Accord – Observations’. It recorded a “significant delay” in deployment into the scene to treat injured casualties.1179

The NWAS debriefs commented that there was a “huge delay” in the deployment of GMFRS and NWAS personnel to triage and treat casualties. It said that the triage teams were only deployed at 02:20. This was 2 hours and 20 minutes after the simulated “attack”.1180 The NWAS debrief said that NWAS and GMFRS personnel should have been able to deploy to the Operation Plato warm zone outside of the building within 30 minutes of the attack. It concluded: “This delay would unequivocally have resulted in unnecessary loss of life.”1181

In its exercise report, NWAS made a recommendation for further joint exercises to include joint agency working. The basis of this recommendation was said to be: “Part of the decision making process could not be made at the operational scene due to one of the agencies missing.”1182 This in turn was said to have led to a time delay in getting to patients.1183

GMP’s view: Forward Command Post

GMP’s view of Exercise Winchester Accord was starkly different to those of GMFRS and NWAS. It developed over the course of the Inquiry.

In its opening statement, GMP accepted that, during Exercise Winchester Accord, the FDO was overstretched and at times impossible to contact. These weaknesses were identified through debriefs after Exercise Winchester Accord. Although GMP said that the exercise was not comparable to real‑life conditions, it accepted that insufficient steps were taken before the Attack to provide extra support for and better access to the FDO.1184 It was agreed that the fact an FDO might be overwhelmed was widely known.1185

When GMP made its closing statement, its analysis was that there was no evidence of a catastrophic failure of the FDO. It submitted that there was evidence that the FDO’s performance during the exercise was very positive.1186 It was also said that there was no delay in deploying NWAS and GMFRS into the Operation Plato warm zone to treat casualties, and there was no JESIP failure detrimental to the wellbeing of casualties.1187 As the FDO was not working from his normal location in Exercise Winchester Accord, it was suggested it was difficult to draw any meaningful parallels with the emergency response to the Attack.1188

A critical reason for the development of the GMP position was a July 2021 statement from Superintendent Openshaw, the Ground Assigned Tactical Firearms Commander for Exercise Winchester Accord on day one.

In his statement, Superintendent Openshaw explained that, in order to inject some reality into the exercise, he waited for a call to activate him as the Ground Assigned Tactical Firearms Commander.1189 He stated that this was supposed to be from the Cadre Tactical Firearms Commander at GMP HQ. However, he explained there was a delay of up to an hour in him receiving that call. He believed this was because the activation of the Ground Assigned Tactical Firearms Commander was not included in the GMP planned sequence of events for the exercise:1190 it had been “overlooked”.1191 He said that he was eventually activated to attend by the Firearms Tactical Advisor, Sergeant Frederick Warburton, and arrived at the Trafford Centre about ten minutes later.1192 On arrival, Superintendent Openshaw stated he had briefings with the Operational Firearms Commander and a military commander in the CCTV control room. This took no more than 15 minutes.1193

Superintendent Openshaw said that he then made his way to the FCP. This took a couple of minutes.1194 As an estimate, based on an arrival time of about 01:30, this would mean that Superintendent Openshaw arrived at the FCP at the earliest between 01:45 and 01:50. The sequence of events for Exercise Winchester Accord indicated that Superintendent Openshaw should have been at the FCP at 01:00.1195

When he arrived at the FCP, Superintendent Openshaw said he could hear gunfire coming from the direction of the Trafford Centre. NWAS and GMFRS commanders were already present. There were no senior GMP officers.1196 Superintendent Openshaw said about the GMFRS and NWAS commanders that he “gained the impression they had been at the FCP waiting for me to arrive”. They had received information from the Silver Control Room that the area outside the Trafford Centre was potentially an Operation Plato warm zone with casualties and wanted to move forward.1197

Superintendent Openshaw said this information was different from the briefing he had received from the Operational Firearms Commander. He believed that Silver Control was “slightly ahead of where the exercise actually was on the ground”.1198 The Operational Firearms Commander had told him that the area immediately around the Orient entrance was an Operation Plato hot zone as the terrorists were positioned so they could fire down into the area. GMP firearms officers were in the Operation Plato hot zone.1199 Superintendent Openshaw shared that information with the NWAS and GMFRS commanders. He stated that this took about 10 to 15 minutes.1200 Taking the shorter estimate, this would take the time to approximately 02:00 or 02:05.

As that briefing finished, Superintendent Openshaw stated that the military moved into the Trafford Centre through the Orient entrance. This allowed GMP firearms officers to move in and sweep the lower floor. A joint assessment of risk was undertaken, and the area outside the Orient entrance was declared an Operation Plato warm zone.1201 The GMFRS and NWAS specialist responders moved in to treat and extract the casualties.1202 NWAS commented in its debrief that its triage teams deployed into the Trafford Centre at 02:20. About 30 minutes later, the terrorists were neutralised and the whole of the Trafford Centre was declared an Operation Plato warm zone.1203

Superintendent Openshaw accepted that his delayed arrival caused “some initial difficulties”.1204 He estimated that it took him about 30 minutes to establish control once he arrived at the Trafford Centre. He said there was a “fractured communications picture” due to “Silver” being ahead in the exercise scenario. Overall, he said the multi‑agency response worked very well.1205 Superintendent Openshaw did not agree that there was a JESIP failure, apart from that caused by his late arrival.1206 Superintendent Openshaw also stated that the delay in him reaching the FCP did not delay GMFRS and NWAS moving forward to casualties. That was because, at the point of his arrival, the Trafford Centre was still an Operation Plato hot zone.1207

What is clear from Superintendent Openshaw’s statement is he was delayed in joining the exercise. This delay was because he was not notified by the Tactical Firearms Commander, and not because of any issue with the FDO. However, Superintendent Openshaw’s statement does not address the timing of NWAS and GMFRS being notified of the Operation Plato declaration or the calling forward of NWAS and GMFRS to the FCP.

In the course of his statement, Superintendent Openshaw referred to a sequence of events for Exercise Winchester Accord. In that document, at 00:09 the FDO was identified as declaring Operation Plato. Under the list of anticipated actions was: “Inform GMFRS and NWAS NILO.” That document anticipated that the FCP would be established by firearms officers at 00:22. At 00:40, it indicated that the FDO should “ensure someone nominates a FCP and informs all necessary staff”.1208

Based on the accounts of those from GMFRS who were involved, these were not communicated to them either at or near the time they were supposed to be.

GMP’s view: Force Duty Officer

CI Booth was the duty officer umpire for the FDO.1209 This role was to ensure that certain actions on the exercise happened at particular points and to ensure that, if the FDO failed to complete a task, he could step in to allow the exercise to continue.1210 If an error was made which needed to be corrected during the exercise, that could be dealt with later through feedback.1211 The FDO for the exercise, Inspector Williams, was “relatively” experienced and a former firearms officer, which meant he was familiar with that aspect of the role.1212

CI Booth said that he could not recall in great detail what happened during the exercise, but that it highlighted that the FDO needed more staff if the move to the Force Hub at GMP HQ was to work.1213

During the exercise, the FDO was based in the Force Command Module, partitioned off from the rest of the Silver Control Room.1214 CI Booth said he was only aware of delays at the RVP and FCP after the exercise. Within the control room, he was not aware of those problems.1215 He was not informed in the planning for the exercise that the FDO was expected to contact the GMFRS and NWAS NILOs within the first nine minutes. CI Booth said that did not “sound an unreasonable element of the exercise”.1216

CI Booth agreed that, in simple terms, the FDO was overwhelmed during the exercise.1217 He was careful to say that he could not recall in great detail what happened during the exercise, but it was seeking to test proposed changes, not the system in place. CI Booth recalled that the FDO coped reasonably well but was frustrated by the inability to delegate tasks.1218 He agreed, in further questioning, with a suggestion that there was a catastrophic failure because there was a failure to declare Operation Plato to GMFRS and NWAS and a failure to call them forward, and that this delay would have likely contributed to a loss of life.1219

CI Booth said he could not recall any issue that was raised about such a significant delay. He felt that he would have stepped in as the exercise would have come to a “grinding halt” and that the exercise co‑ordinators would have corrected the situation.1220 It was only after the exercise he became aware of such a huge delay.1221 He accepted that the failures in Exercise Winchester Accord were very similar to those that occurred 12 months later in the response to the Attack.1222 This evidence was provided by CI Booth in March 2021, four months before the statement of Superintendent Openshaw was provided to the Inquiry.

Inspector Roby described herself as a “roving problem solver” based at GMP HQ.1223 As she had written the majority of the exercise, she was the overall co‑ordinator. Umpires were allocated to each of the individual areas, so they were close by to deal with anything that went wrong and note down good and bad points.1224

Inspector Roby said that, as she was walking around, somebody told her that they had not received an activation call from the FDO. While it was her impression that the FDO “managed extremely well under the circumstances”, she agreed that this did not in fact appear to be the case.1225 She understood that communication between agencies went wrong at some point during the exercise, and this identified a need for more JESIP training.1226

Inspector Roby said that she only learned later that there was a lack of communication at the FCP and problems with the Ground Assigned Tactical Firearms Commander going there.1227 She said that, if she had known, she would have got involved to help sort it out. It was a major part of the exercise, Inspector Roby said, but: “Unfortunately, the ODU [Operational Development Unit] were more interested in military relicensing than they were in multi- agency exercising.”1228 Inspector Roby was not the only witness to refer to such issues. The tension between the different objectives for different organisations from the exercise was a recurring theme. It was inevitable that this fed into tension between which objectives had the greater priority and the overall efficacy of the exercise.

CI Booth explained that an important lesson he took from the exercise was that the proposed move of the FDO to the Force Hub at GMP HQ would not work. He also accepted that the fact that the Exercise Winchester Accord arrangements had failed did not necessarily mean that the existing provision for the FDO, with 30 or 40 staff, would work. The outcome of Exercise Winchester Accord was not, as CI Booth stated, “an automatic pass” for the existing arrangements.1229 He agreed that there was no guarantee that more people would make it any better.1230 However, he explained that it was impossible to replicate the control room in a real‑life scenario as they needed to keep working 24 hours a day. To do so would leave GMP “dangerously vulnerable”.1231 CI Booth accepted that it was less than ideal that there was no way of testing whether the FDO was going to succeed with the existing system.1232

Inspector Roby said that, at the time, she was not aware that there had been such a big delay in deploying ambulance and fire and rescue personnel into the Operation Plato warm zone. She said that she “agree[d] entirely” with the debrief comments made by NWAS about them not being deployed into the warm zone within 30 minutes and the delay potentially leading to an unnecessary loss of life.1233 She agreed that was a learning point from the exercise.1234 There were no surprises in the outcome of the exercise for Inspector Roby. It established that the FDO could not work in isolation.1235 The risk of the FDO being overwhelmed would only increase by moving the FDO to GMP HQ.1236

As a result of Exercise Winchester Accord, a review was undertaken of the proposed move of the FDO to GMP HQ. The move was only considered feasible once the remainder of the Operational Communications Branch had also transferred across. On the night of the Attack, the FDO was still located in the Operational Control Room.1237 Inspector Roby said that she was not aware of the actual set‑up of the FDO in the control room being tested before the Attack.1238

It is telling that the outcomes of Exercise Winchester Accord told Inspector Roby exactly what she expected to hear. Her focus and that of her GMP colleagues was on the Force Hub proposal. It is unfortunate that the acknowledged problems that the FDO experienced on Exercise Winchester Accord did not flag the need for more testing of the existing arrangements of the FDO working from the Operational Control Room.

There were problems that arose during the exercise with the capacity of the FDO and joint working with NWAS and GMFRS. The extent of those problems needed to be more thoroughly understood and could not simply be explained by testing the FDO role from the Force Hub at GMP HQ. The failure to conduct a more critical and searching analysis of the lessons from Exercise Winchester Accord will be considered in the following section on debriefs from the exercise.

Debriefs following Exercise Winchester Accord

Joint Organisational Learning was introduced as part of JESIP in 2015. This mechanism was introduced because the emergency services were frequently identifying issues, but they were not being shared nationally for wider learning or leading to changes in local practice.1239 JESIP promoted a framework of hot debriefs led by commanders immediately after an event and formal, structured debriefs co‑ordinated by lead agencies.1240 The Policing Experts said that debriefs should “capture aspects that were positive alongside those aspects that did not go so well”.1241 The debrief process on Exercise Winchester Accord fell below these aspirations.

On 11th May 2016, CI Booth received an email from an Operational Communications Branch Trainer, Kelly Chilton, with feedback on the performance of call handlers. She said: “The stress levels in the room were intense.”1242 She said that call handlers needed a reference document to help them to know what to do during a Marauding Terrorist Firearms Attack.1243 CI Booth agreed that this feedback appeared to indicate that the action cards were not tested by call handlers during Exercise Winchester Accord.1244

On 13th May 2016, Group Manager Fletcher emailed GMP Superintendent Giladi about Exercise Winchester Accord. Joe Barrett from NWAS was on copy.1245 The email identified that an issue from the exercise was “the linkage in particularly [sic] in the initial stages of the incident, with the Tactical Firearms Commander and the GMFRS/NWAS Commanders”.1246 It said that only a marshalling officer was deployed to the initial RVP and linkage with the Tactical Firearms Commander was very limited. This led to an “excessive delay” in GMFRS and NWAS resources moving forward. The email said this delay was approximately two and a half hours.1247 Group Manager Fletcher later stated that it was a “slip of the keyboard” and the email should have said one and a half hours.1248 He suggested arranging a one‑day joint awareness course for commanders.1249

Inspector Williams emailed CI Booth on 14th May 2016 with his “[t]houghts re Ex Winchester Accord”.1250 He said, “you have to know your limits … what you can achieve before you become overloaded”.1251 The email continued that there was a need to prioritise actions and get support to the FDO as soon as possible. He said that “the reality is the FDO will be frazzled”.1252 CI Booth accepted that this was a reference to the overloading and overwhelming of the FDO.1253

Inspector Williams provided further feedback in his email that the process would run more smoothly if staff had an understanding of Operation Plato, the use of language and what a Marauding Terrorist Firearms Attack response looked like.1254 CI Booth agreed that this was a further indication that the draft action cards he had prepared were not used or tested during the exercise.1255

Inspector Williams also commented that a lot of people will think they are the most important unit but “the truth is, it is all about getting our guns down there”.1256 When setting out his priorities, Inspector Williams listed “getting ARVs [Armed Response Vehicles] to the scene” first, “[t]hen we can start to look to mobilise the Fire and HART, although they are likely to have already heard and been in contact”.1257

From his perspective, CI Booth considered that the fire and rescue and ambulance role in casualty treatment needed to be deployed as soon as they were able.1258 He agreed that “it certainly appears” that there was a preoccupation with the deployment of firearms officers to the detriment of deploying the fire and rescue and ambulance services.1259 This indicated a lack of understanding of the need for a multi‑agency response to a Marauding Terrorist Firearms Attack. Operation Plato is far more than an armed response, and this comment should have been a flag that more training was needed for the FDO role in JESIP and multi‑agency working.

The GMP structured debrief took place on 16th May 2016.1260 The debrief team was one GMP officer, PC Hughes, and two GMFRS officers. It was attended by 14 key personnel from the Strategic Co‑ordination Centre who were activated during the exercise, including Superintendent Openshaw. Inspector Williams did not attend, but a questionnaire he completed was read out.1261 There was no evidence whether the email feedback provided by Inspector Williams was also available, but it seems unlikely as CI Booth did not participate in the structured debrief.

The GMP structured debrief report identified over 50 areas for improvement. Feedback was provided on the “[c]onfusion between the role of TFC [Tactical Firearms Commander] in Silver and ground TFC” and that there was “[n]ot enough information at the FCP as to what was happening inside the building”.1262 There was also a comment that the Tactical Firearms Commander was called in too late and “always playing catch up” as the military assets had already deployed before the Ground Assigned Tactical Firearms Commander was in place.1263 Fifteen areas that went well were identified, including: “The FDO did a great job, was knowledgeable and knew what to do” and “JESIP worked well, three blue lights speaking the same language at the FCP in the warm zone. No C&C [command and control] issues.1264

GMP debrief questionnaires

Seventeen questionnaires were prepared for the GMP structured debrief.1265 Participants were asked to say what aspects of the Strategic Co‑ordination Centre did not go well and what recommendations they would make. The questionnaires included references to the Ground Assigned Tactical Firearms Commander being assigned too late, the need for a better understanding of the acronyms used during the deployment and the need to keep better track of who had been contacted. The questionnaires also identified the need for more clarity around the communication by the FDO on the command and control structure for the response.

The questionnaire from a Ground Assigned Tactical Firearms Commander, CI Sarah Morton, who shadowed Superintendent Openshaw,1266 referred to “confusion” between the role of the Tactical Firearms Commander in the Silver Control Room and the Ground Assigned Tactical Firearms Commander and “management of partners at FCP”. She suggested there was “someone at FCP to co-ordinate and communicate with partners” but that the “response from partners on the ground was good”.1267

The questionnaire from Superintendent Jim Liggett, the Tactical/Silver Commander for the exercise, noted that “contact from FDO came late in the day” and a “significant amount of activation had already taken place (66 pages of FWIN [Force Wide Incident Number])”.1268 Superintendent Liggett queried the definition of the Operation Plato warm zone if there was an Improvised Explosive Device (IED) on a body. This issue should have been resolved as part of the debrief. This might have forestalled some of the issues on 22nd May 2017. Superintendent Liggett linked this comment to an observation about “managing expectations of NWAS and GMFRS to attend the Orient before declared a ‘warm zone’”.1269

CI Lisa Wroe, the Tactical Firearms Commander in the Silver Control Room, commented: “[I]mmediate actions completed by FDO, however slow to contact. TFC [Tactical Firearms Commander] – I had to contact.”1270 She commented that it was not obvious that she had taken command and control. It was communicated “on air” but not picked up by the FDO so she had to interject.1271

CI Wroe commented: “Ambulance informed at 00:01 that they could enter warm zone, however clearly not communicated properly as they asked later on.” She said that she was “[a]sked countless times / pressure re warm zones when the venue was not safe.”1272 CI Wroe concluded that it was a “great learning exercise”, it was “invaluable for interoperability and partners” and it highlighted the “complexities of such an incident”.1273

GMP debrief recommendations

Recommendations arising out of the GMP debrief were allocated to named individuals. The 19 recommendations listed included: JESIP training needed for officers and commanders; the need for a police liaison at the FCP to co‑ordinate and communicate with partners; and additional support for the FDO would be necessary in the event of a real incident.1274 There was no evidence to indicate that such additional support was provided.

The recommendation relating to the support for the FDO was assigned to CI Booth.1275 CI Booth explained that he thought the support already in place for the FDO was sufficient to discharge this recommendation, and no action was taken. In hindsight, he agreed that assessment was mistaken.1276 There was no evidence about steps taken to implement other recommendations about the police liaison at the FCP and JESIP training.

The GMP structured debrief, questionnaires and email observations were collated within a week of Exercise Winchester Accord.1277 This was commendable, but there were significant gaps in the learning captured on the structured debrief.1278

Nothing appears to have been captured about the emailed feedback from the FDO to “know your limits”, and that the FDO would be “frazzled” and “overloaded”.1279 Nothing was done to capture his observations on prioritising the deployment of the armed response. Observations from control room staff about the need for a reference document were not highlighted on the structured debrief report. There was also nothing to indicate an analysis of the 62 local objectives set for GMP against the areas for improvement and recommendations.

On the information available, the GMP structured debrief lacked focus. It failed to scrutinise properly problems that arose during Exercise Winchester Accord. This represented a significant missed opportunity to capture learning from the exercise. If GMP had done so, the problems that did arise with the FDO, whether or not catastrophic, and the issues that arose at the FCP would have been captured and understood much better.

Greater Manchester Resilience Forum multi-agency debrief

A GMRF multi‑agency structured debrief took place on 23rd May 2016. Representatives from GMFRS, NWAS and GMP were present, together with other agencies involved in the exercise. The structured debrief report identified over 40 areas for improvement.1280

Six participants in the GMRF debrief commented they had been “informed late into the incident”1281 and two participants said they “[n]ever received a call out”.1282 Three participants said that an effective assessment of risk was hampered by a “lack of tri-service commanders coming together” and poor communication at the scene.1283 GMFRS was not aware of the set‑up of the warm zone, and it was noted: “If the FDO is busy the person who answers the phone may not know what to do, we need a different route to the FDO for the setting up of the SCC [Strategic Co-ordination Centre].”1284

The GMRF debrief report identified 25 areas that went well, including good multi‑agency partner working. The FDO and Tactical/Silver Commander were praised.1285 Nineteen recommendations were made. None of the recommendations focused on the role of the FDO or JESIP working at the FCP. One recommendation commented on the need to ensure key personnel were present at exercises, such as IT, media and BTP. No reference was made to NWFC. Only four of the recommendations had named individual owners to implement them.1286

It is difficult to be confident about how robust the analysis conducted for the GMRF debrief was. There did not appear to have been a check back against all the local objectives or consistent ownership of the recommendations. It appears that the GMRF debrief was conducted at least before the final NWAS debrief report. I do not have access to a structured debrief report from GMFRS to understand when its debrief took place, if at all.

NWAS provided the Inquiry with “debrief points” from the exercise.1287 These were undated and have previously been summarised to set out the NWAS view of the exercise. The notes included two very short positive comments: “good triage by AIT team [Ambulance Intervention Team]” and “positive attitude by team”. The notes identified 17 “negatives”. These included reference to a “huge delay” to having a tri‑service meeting, and delays to deployment of NWAS and GMFRS to triage and treat. It was said that there was a lack of direct police on‑site liaison with NWAS and GMFRS. The negatives also included issues that arose between GMFRS and NWAS in respect of co‑ordination and leadership around triage, treatment and recovery of casualties.1288

Following the GMRF structured debrief, the next significant staging post in the debrief process from Exercise Winchester Accord was a meeting of the Blue Light Forum on 28th June 2016. This was attended by representatives from GMP, GMFRS and NWAS.1289 The minutes from the Forum noted that “[t]he main issue was the delay in getting NWAS and Fire Service resources to the incident scene”.1290 It was said that this resulted in a two‑hour delay in deployment. The possibility of further awareness training was discussed for Tactical Firearms Commanders.1291

The following day, a meeting was arranged between Superintendent Giladi, Joe Barrett from NWAS and Group Manager Fletcher. This was to discuss Group Manager Fletcher’s email sent shortly after the exercise on 13th May 2016.1292 Superintendent Giladi said he engaged positively with the email because he was concerned about what was being raised and had a good relationship with GMFRS and NWAS.1293 Superintendent Giladi’s daybook recorded a brief note of the meeting.1294 A note with an asterisk read: “Co-location!! – same mistake every time.1295 He said that this point was “clearly of concern” and that it had been an issue during several exercises.1296

In his evidence, Superintendent Giladi stated that he had understood from Group Manager Fletcher that JESIP probably was not applied during Exercise Winchester Accord. He understood that, “there was certainly what appeared to be a lack of communication on the ground to ensure that Fire and Rescue and Ambulance Service resources were used to their best ability”.1297 Superintendent Giladi said that such an excessive delay would have had potentially “horrendous consequences”.1298 He said that it was agreed to set up training on command and control. This became the JOPs commander training that took place in January and February 2017.1299

The NWCTU debrief report from Exercise Winchester Accord was finalised on 5th July 2016.1300 The report provided important insight into what happened during the exercise at the FCP.

Dealing with the designation of the FCP, the report noted that, when military assets arrived, there was no Ground Assigned Tactical Firearms Commander. As a result, the Operational Firearms Commander performed the role. The Operational Firearms Commander designated the FCP and briefed the military to agree a tactical plan.1301 It was said that, at this stage, the FDO handed over command to the Cadre Tactical Firearms Commander, but they were uncontactable.1302

On the delayed declaration of the Operation Plato warm zone, it was said that the Tactical Firearms Commander was unsighted on the process of neutralising “subjects” and clearing areas. Therefore, the Tactical Firearms Commander was not confident in declaring the warm zone until the Operational Firearms Commander had provided tactical advice to her.1303 At this point, “JESIP partners deployed in a casualty management role”.1304 These are important points. They are not reflected in the GMP or GMRF structured debriefs. This reinforces the concern about the quality and consistency of the debrief process and learning lessons.

On 7th July 2016, Superintendent Giladi chaired GMP’s Major Incident Public Order and Events Group meeting.1305 One of the attendees was Superintendent Openshaw. An update was given on the recommendations from the GMP structured debrief. It was highlighted that, “[t]he exercise had been run specifically to test whether the FDO could work in isolation and it had been established that this did not work”.1306 Despite the meeting that Superintendent Giladi had with Group Manager Fletcher only the week before, it seems unlikely that the issues relating to co‑location were referred to at that meeting. It would have been helpful if this was discussed, particularly with Superintendent Openshaw present at the meeting, to embed learning on this critical issue with all the emergency services.

On 21st July 2016, there was a meeting of the GMRF Resilience Development Group.1307 This was attended by Superintendent Giladi and Inspector Roby for GMP, Group Manager Fletcher and Station Manager Berry for GMFRS, and two representatives for NWAS. The NWCTU debrief report on Exercise Winchester Accord was presented to the meeting.1308 It is not apparent that the problems of co‑location were raised at the meeting.

The overall debrief process identified many learning points from Exercise Winchester Accord. However, it was disjointed and lacked the rigour that is necessary to track each exercise objectively against lessons learned and recommendations. Too often, opportunities were missed to reflect on issues that arose during Exercise Winchester Accord because the importance of the information was not understood or was simply not shared with all agencies. This applies to the debriefs conducted by all the emergency services and GMRF. Improvements must be made to debrief properly from large exercises and to ensure an appropriate level of resource is provided to achieve this.

The Policing Experts recommended that local resilience forums “should be more closely involved in managing the lessons to be learned from major exercises, or serious incidents, in their areas and for the specific debriefing of those events”. They noted that local resilience forums currently have “no audit or assessment mandate to ensure that multi-agency arrangements are effective or are supported adequately by single agency plans or capability”.1309 This is a sensible recommendation, and consideration should be given by central government as to how to make the debrief process more effective.

Exercise Winchester Accord conclusions

Exercise Winchester Accord was an ambitious exercise.1310 It offered an important opportunity to conduct a live exercise of a Marauding Terrorist Firearms Attack scenario. Ultimately, the exercise tried to do too much. There were too many local objectives and, without the funding and support available for an equivalent national exercise, it was inevitable that Exercise Winchester Accord could not deliver on all its objectives.

Important learning was identified from the exercise. It was apparent that the proposed move of the FDO to the Force Hub at GMP HQ would not work.1311 Yet Exercise Winchester Accord cannot be regarded as a success. It foreshadowed some of the problems in the emergency response that were to arise on 22nd May 2017. Although the role of the FDO was tested in a different location from where they worked on the night of the Attack, parallels can still be drawn between the failures in Exercise Winchester Accord and problems that arose a year later at the Victoria Exchange Complex.

The FDO’s performance cannot be described as a catastrophic failure. There is, however, evidence that he was overwhelmed during the exercise, particularly the early stages. This contributed to the failures to communicate the Operation Plato declaration and the delays in deployment to the FCP. These were problems which are relevant to what was to go wrong on 22nd May 2017.

Based on Superintendent Openshaw’s statement, the reason why NWAS and GMFRS thought they had been delayed in reaching casualties was because he was not deployed to the FCP when he should have been. In its closing statement, GMP recognised why that perception would have arisen at the time the exercise was under way.1312

I find it remarkable that this did not emerge during the debrief process in a way that meant all participants were aware of it. This is particularly so given the concern about delay that NWAS and GMFRS had at the time. An effective debrief process would have resolved this. It had a direct bearing on what conclusions could, and could not, be drawn.

Agreement as to what went wrong and why should have been reached at the time. A shared understanding of what took place is vital to the process of making improvements. Once there was agreement about what occurred, a constructive discussion should have taken place in relation to the learning that was to be derived. That would have formed a platform for positive change in this difficult and important area of an emergency response. As it was, those involved went their separate ways holding different views about what needed to change.

Each organisation was focused on its own objectives for the exercise. Everyone had put a great deal of preparation into its organisation. It seems likely that this exaggerated failings when individual aspects of the exercise did not play out as anticipated. GMFRS, in particular, felt aggrieved by the failures at joint working. GMP officers were satisfied that the exercise had proved what they already knew, that moving the FDO to a Force Hub would not work, but failed to look at other reasons why the FDO was overwhelmed during the exercise.

More generally, the debrief process on the exercise was inadequate. It failed to track objectives against what happened during the exercise and identify consistent lessons.1313 The process of structured debriefs was therefore not robust and did not offer a forum to identify the systemic problems which were repeated in the Attack 12 months later.

Exercise Winchester Accord represented a significant missed opportunity to prepare an adequate and robust response to a Marauding Terrorist Firearms Attack, or similar incident, within Greater Manchester.