Description
Incident management (IcM) is a term describing the activities of an organization to identify, analyze, and correct hazards to prevent a future re-occurrence. These incidents within a structured organization are normally dealt with by either an Incident Response Team (IRT), or an Incident Management Team (IMT).
CFRS : Incident Response Modelling Project
Incident Response Modelling Project
Cheshire Fire and Rescue Service (CFRS).
Richard Smart Process Evolution Limited Version 1.3 17th November 2006
Tel: 07782 170337 Fax: 08701 709887 E-mail: [email protected] Web: www.processevolution.co.uk Process Evolution Limited, 6 Fordbank Court, 218 Henwick Road, Worcester. WR2 5PF Registered in England No. 4983030, VAT 823-9689-84
CFRS : Incident Response Modelling Project
Executive Summary
This document forms the Interim Report for Cheshire Fire & Rescue Service (CFRS) as part of the Incident Response Modelling Project being conducted by Process Evolution on their behalf. This report is a key deliverable for the “Diagnose” phase of the project. It contains the results of initial analysis of data captured from CFRS operational systems and has two main purposes: 1) 2) To ensure an analytical understanding of CFRS operations and data prior to commencing the “Evaluate” phase of the project. To provide CFRS with detailed analytical insight into a range of proposed modelling scenarios prior to evaluation.
The report provides a high level analysis of Incident Volumes and Operational Performance of CFRS based on Incident and Deployment data between 1st January 2003 and 30th September 2006. It also provides initial insight into 5 CFRS specified modelling scenarios which will be the subject of further simulation assessment in subsequent stages of this project: Birchwood Fire Station Wilmslow Fire Station Northwich / Winsford / Middlewich Fire Station Reconfiguration Targetted Response Vehicles (TRVs) Deployment Aerial / Hydraulic Lift Platform Deployment
Summary conclusions to be drawn from this analysis are: 1 High volume of incidents requiring attendance are false alarms (over 40%). The Incident Response Simulator could be used to look at the effects of investigative attendance strategies on resource loading. No clear trend in overall incident numbers, although they appear to be reducing. There is clear seasonality in incident volumes. The Incident Response Simulator could be used to investigate a range of appliance rostering solutions. Significant variance in overall daily incident workload (Small Fires in particular). This suggests greater use could be made of TRV’s during peak periods. Emergency Response Performance CFRS are achieving a 98% (3½ year average) Emergency Response performance against the 2006/07 Emergency Response standards at Brigade level. No discernable difference between 8 vs 9 Rider attendances on Incident Times.
2 3 4 5
6
These conclusions are discussed in greater detail in the body of the report. The next steps are to complete the ‘Evaluate’ Phase as follows: - Configure the Incident Response Simulator to use CFRS Incident Data. - Calibrate the Simulator against known CFRS performance. - Design an experimentation plan to evaluate the five scenarios identified above.
Version : 1.3
Page 2 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Table of Contents
Executive Summary .......................................................................................................... 2 Table of Contents ................................................................................................................ 3 Document History ................................................................................................................ 3 1 2 3 4
4.1 4.2 4.3 4.4 4.5 4.6
Introduction................................................................................................................. 4 Background................................................................................................................. 5 Approach..................................................................................................................... 7 Analysis – Overall Incident Volumes ........................................................................ 8
Incidents by Type.................................................................................................................................... 8 Incidents by Year .................................................................................................................................. 10 Incidents by Month................................................................................................................................ 12 Incidents by Day of Week ..................................................................................................................... 13 Incidents by Hour of Day....................................................................................................................... 14 Incidents by Station .............................................................................................................................. 15
5
5.1 5.2 5.3 5.4
Analysis – Overall Performance .............................................................................. 16
Response Times................................................................................................................................... 17 Response Delivery................................................................................................................................ 18 Resource Deployment .......................................................................................................................... 20 Casualties............................................................................................................................................. 21
6 7 8 9
Analysis - Scenario Assessment............................................................................. 22 Analysis – Abstraction Data .................................................................................... 32 Analysis – 8 Riders vs 9 Riders............................................................................... 33 Conclusions and Next Steps ................................................................................... 34
9.1 Conclusions .......................................................................................................................................... 34 9.2 Next Steps. ........................................................................................................................................... 34
Appendix 1 – Glossary of Terms.................................................................................... 35
Document History
Version
0.1 1.0 1.1 1.2 1.3
Release
Initial Draft Interim Report Final Release Candidate (Post completion of IAT & Risk Planes) Final Release Candidate Final Release Final Release
Date
10 October 2006 2nd November 2006 5th November 2006 14th November 2006 17th November 2006
th
Version : 1.3
Page 3 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
1
Introduction
This document forms the Interim Report for Cheshire Fire & Rescue Service (CFRS) as part of the Incident Response Modelling Project being conducted by Process Evolution on their behalf. This report is a key deliverable for the “Diagnose” phase of the project. It contains the results of initial analysis of data captured from CFRS’s operational systems and has two main purposes: 1) 2) To ensure an analytical understanding of CFRS operations and data prior to commencing the “Evaluate” phase of the project. To provide CFRS with detailed analytical insight into a range of proposed modelling scenarios prior to evaluation.
Version : 1.3
Page 4 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
2
Background
The Government’s National Framework Strategy challenges fire services to review the deployment of their resources in order to provide targeted response in a more cost effective manner. Previous incident response standards and corresponding deployment strategies are based on historical guidelines and risks. As part of its response to these challenges CFRS is considering a number of changes to its deployment strategy for emergency response, such as the: • • • Mix between whole time (WT), retained (RT) and day-crewed (DC) appliances. Identifying optimal station bases for specialist appliances. Crewing/use of appropriate response vehicles such as Targeted Response Vehicles.
Process Evolution specialises in the design, analysis and continuous improvement of business processes. A key feature of our work is the use of quantitative techniques which enable the impact of proposed changes to be quantified prior to implementation. CFRS have therefore commissioned Process Evolution to undertake initial evaluations of specific proposed changes, and to help the service establish an ‘in house’ capability to evaluate such changes in the future. The specific proposed changes to be evaluated are: Birchwood Fire Station Quantify the impact of providing cover to the Birchwood area from Warrington at certain times of day or changing availability from Whole Time to Day-crewed, or other appropriate response options. Wilmslow Fire Station Evaluate the effect of changing availability from Whole Time to Day Crewing / Retained, or other appropriate response options. Northwich / Winsford / Middlewich Fire Station Reconfiguration Evaluate the effects of changing appliance availability mix across these stations. In particular, it may be possible to provide a reduced level of immediately available cover at certain times, with a 1 hour response say for additional appliances in the event of a major incident, or other appropriate response options. Targetted Response Vehicles (TRVs) Deployment Identify the station bases and operating model for TRV’s and corresponding hours worked that will deliver optimal response times across CFRS. Aerial / Hydraulic Lift Platform Deployment Identify the station bases for HP’s that will deliver optimal response times across CFRS. Look at effect of reducing availability of one HP to retained.
For performance purposes CFRS is structured into 8 Performance Delivery Groups (PDGs) one for each of its Local Government Districts and Unitary Authorities. These are individually colour coded on the map on the following page.
Version : 1.3
Page 5 of 35
Print Date : 17/11/2006
Version : 1.3
Performance Delivery Groups
Chester Congleton Crewe and Nantwich Ellesmere Port and Neston Halton Macclesfield Vale Royal Warrington
CFRS Station Map
CFRS : Incident Response Modelling Project
Page 6 of 35
PDG
Figure 2.1
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
3
Approach
The diagram below shows the main steps in the project.
Consulting Approach
Monitor performance
Agree Proposal
Agree Project Scope
Diagnose
Evaluate
Plan
Handover
Provide Support
Deliverables Terms of Reference Interim Report Model Runs Report Final Report
Analysis Toolkit
Figure 3.1
The “Scoping” phase resulted in the identification of the five scenarios detailed for initial evaluation and summarised in the previous section. This document reports on the ‘Diagnose’ phase of the project. During this phase, we extracted information from CFRS’s command and control system relating to incidents occurring in the calendar years 2003 – 2006, providing 3 ½ years of historical information referred to as the analysis period throughout this document. This forms a statistically robust and representative sample of data on which to undertake the analysis for this project. We have configured our Incident Analysis Toolkit (IAT) to import this data and have used it to provide the analysis contained in Section 4. Additional discussions have been held with CFRS staff in order to further understand the deployment process. This has been incorporated in the Simulator. Work has already begun on configuring our Incident Response Simulator to reflect the infrastructure of CFRS in terms of the number and location of fire stations and the respective appliances based at each one. During the ‘Evaluate’ phase, we shall use this model to test the proposed changes and quantify the impact on response performance against CFRS’s response standards for first attendance at incidents. These will be based on the 2006/07 Response Standards published by CFRS for each Risk Plane associated with a Station.
Version : 1.3
Page 7 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4
Analysis – Overall Incident Volumes
This section contains results from analysis of CFRS incident data using the Incident Analyser Tool (IAT). Unless otherwise stated, the graphs relate to incidents occurring in the period 1st January 2003 – 30th August 2006 which is referred to as the analysis period.
4.1
Incidents by Type
The chart shows the total number of incidents of each specific type by PDG Area that occurred during the analysis period:
Incident Volumes By Specific Incident For All Years
25000
W arrington
Vale R oyal
20000
M acclesfield
Incident Volum
H alton
15000
E llesm ere P ort and N eston
10000
C rewe and N antwich
5000
C ongleton
0 146 - SSC 142 - FDR1 144 - FAGI 143 - FAM 149 - Chimney Fires 147 - Small Fires
C hester
Specific Incident
Figure 4.1.1 It can be seen that the largest category is Small Fires (147) closely followed by False Alarms Good Intent (144).
Incident Volumes By PDG For All Years
14000 12000
149 - C him ney Fires
147 - Sm all Fires
Incident Volum
10000 8000 6000 4000 2000 0 Macclesfield Halton Ellesmere Port and Neston Crewe and Nantwich Warrington Vale Royal Congleton Chester
143 - FAM 146 - SSC
144 - FAG I
142 - FD R 1
PDG
Figure 4.1.2 Volumes by PDG vary considerably with Warrington and Halton being almost 60% higher than other districts. Types of incident (by %) also vary considerably by PDG as can be seen in the following chart.
Version : 1.3
Page 8 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Incident Volumes By PDG For All Years
100% 90% 80%
149 - C him ney Fires
Incident Volum
147 - Sm all Fires
70% 60% 50% 40% 30% 20% 10% 0% Macclesfield Halton Ellesmere Port and Neston Crewe and Nantwich W arrington Vale Royal Congleton Chester
142 - FD R 1 143 - FAM 144 - FAGI 146 - SSC
PDG
Figure 4.1.3 Small Fires pre-occupy the urban areas of Halton, Warrington, Ellesmere Port and Vale Royal, where as the more rural areas deal with more False Alarm Good Intents (144) and Special Services (146). The greater proportion of “Over the Border” incidents tend to be Special Services (146)
Version : 1.3
Page 9 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.2
Incidents by Year
The next chart shows the number of incidents by PDG over the analysis period:
Incident Volumes By PDG
20000
W arrington
18000 16000
Vale R oyal
Incident Volum
14000 12000 10000
M acclesfield
H alton
E llesm ere P ort and N eston
8000 6000 4000 2000
C hester C rewe and N antwich
C ongleton
0 2003 2004 2005 2006
Period
Figure 4.2.1 There is no clear trend in incident volumes over time however there is a suggestion from the 2003/04/05 that volumes are decreasing. This cannot be determined until a complete set of incident statistics are available for 2006.
Incident Volumes By PDG
4000 3500 3000
W arrington V ale Royal M acclesfield
Incident Volum
2500
Halton
2000
Ellesmere Port and Neston
1500
Crewe and Nantwich
1000 500 0 2003 2004 2005 2006
Congleton Chester
Period
Figure 4.2.2
Version : 1.3
Page 10 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
If we examine this data by incident type, we can see from the chart below (Figure 4.2.3) that most of the decrease from 2003 to 2004/05 is accounted for by a drop in Small Fires (147).
Incident Volumes By Specific Incident
149 - Chimney Fires
20000 18000 16000
147 - Small Fires
Incident Volumes
14000 12000 10000 8000 6000 4000 2000 0 2003 2004 2005 2006
142 - FDR1 143 - FAM 144 - FAGI 146 - SSC
Period
Figure 4.2.3
Version : 1.3
Page 11 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.3
Incidents by Month
The next chart suggests there is some seasonality in the number of incidents – however with an incomplete 2006 data set this is difficult to conclude. The busiest months are July and August, with Small Fires (147), SSC’s (146) and False Alarm Good Intents (144) being the main contributors to these rises in activity levels..
Incident Volumes By Specific Incident For All Years
7000 6000
149 - Chimney Fires
147 - Small Fires
Comment [PL1]: Suggest that the effect of incomplete 2006 is removed by pro-rata or eliminating 2006 data – do this throughout
146 - SSC
Incident Volumes
5000
144 - FAGI
4000 3000 2000
142 - FDR1 143 - FAM
1000 0 September November June January March July October December April May February August
Period
Figure 4.3.1 Incident types vary considerably by month. Chimney Fires (149) only show during the winter months (Nov-Apr) and False Alarms Malicious Intent (143) appear to be proportionately higher during those months. Incident volumes appear to be around 40% higher during the summer months, though there is some volatility around holiday periods.
Incident Volumes By Specific Incident For All Years
100%
149 - Chimney Fires
147 - Small Fires
80%
146 - SSC
Incident Volumes
60%
144 - FAGI
40%
143 - FAM
20%
142 - FDR1
0% September November June January July October December March April February August May
Period
Figure 4.3.2 It is no surprise that Chimney fires occur primarily in the winter months (Nov-Apr) December to March whereas Small Fires (147) tend to be more frequent in the summer months.
Version : 1.3
Page 12 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.4
Incidents by Day of Week
Weekends appear marginally busier than other days of the week:
Incident Volumes By PDG For All Years
10000 9000 8000
W arrington
Vale R oyal
Incident Volum
7000 6000 5000 4000 3000 2000 1000 0 Thursday Tuesday Friday Saturday Monday W ednesday Sunday
M acclesfield
H alton
E llesm ere P ort and N eston
C rewe and N antwich
C ongleton
C hester
Period
Figure 4.4.1 However, volumes on the busiest day (Saturday) are less than 10% more than on the quietest day, which does not in itself suggest the need for a widely different deployment capability on different days of the week. Looking at the proportions of Incident Type by day reveals a subtly different pattern of behaviour. rds There appears to be a 1/3 reduction in the proportion of False Alarms Good Intent (144) at weekends, whereas there is a 20% increase in Small Fires (147) over the same period.
Incident Volumes By Specific Incident For All Years
149 - C him ney Fires
100% 90%
147 - Sm all Fires
80%
Incident Volum
70% 60% 50% 40% 30% 20% 10% 0% Tuesday Thursday Friday Saturday Monday W ednesday Sunday
142 - FD R 1 143 - FAM 144 - FAG I 146 - SSC
Period
Figure 4.4.2 This suggests a need to hold greater Small Fire capacity for weekend cover.
Version : 1.3
Page 13 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.5
Incidents by Hour of Day
The next chart shows when incidents occur through the day over the analysis period:
Incident Volumes By Generic Incident For All Years
5000 4500 4000
149 - C him ney Fires
Incident Volum
3500 3000 2500 2000 1500 1000 500 0 00:00 to 01:00 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
147 - Sm all Fires
146 - SSC
144 - FAG I
143 - FAM
142 - FD R 1
Period
Figure 4.5.1 Here the variance over time is much more striking, with peak demand between 17:00 and 21:00 over six times that between 05:00-06:00. It can also be seen that different incident types have different peak periods. False Alarms Good rd Intent (144) primarily occur from 08:00-22:00 and make up over 2/3 s of incidents between 09:00-10:00, whereas False Alarms Malicious Intent (143) peak between 18:00 and 23:00. Small Fires grow progressively between 12:00 and 24:00 peaking around 19:00-21:00 and account for over 50% of calls at that time of day.
Incident Volumes By Specific Incident For All Years
600 500
Incident Volumes
400 300 200 100 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
142 - FDR1
Period
Figure 4.5.2 This chart shows that Primary Fires FDR1 (142) peak between 20:00 and 23:00 at around 3½ times that of the lowest demand between 05:00-08:00.
This variation in incident demand suggests that some benefit may be derived from looking at appliance shift patterns and matching them more closely to the daily demand profiles illustrated in these two charts.
Version : 1.3
Page 14 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.6
Incidents by Station
The next chart shows the proportion of incidents that occur at each station, together with colour coding to indicate the crewing status of each station. Whole-time Stations cover the highest volume of incidents, though Birchwood and Wilmslow experience lower demand than the Day-crewed Northwich and Winsford Stations.
Incident Volumes By Station Ground For All Years
8000 7000
6000 5000 4000 3000 2000 1000 0 Over the Border Macclesfield Knutsford Stockton Heath Tarporley Chester Winsford Northwich Sandbach Nantwich Poynton Runcorn Crewe Congleton Birchwood Middlewich Bollington Holmes Chapel Frodsham Ellesmere Port Warrington Wilmslow Unknown Audlem Widnes Malpas
Warrington Ellesmere Port Chester Runcorn Crew e Widnes Macclesfield Northw ich Winsford Birchw ood Wilmslow Knutsford Stockton Heath Congleton Nantw ich Frodsham Sandbach Tarporley Poynton Middlew ich Bollington Holmes Chapel Malpas Audlem Over the Border Unknow n
Incident Volumes
Station Ground
Figure 4.6.1
Station Key Whole-time Day-crewed Retained
Version : 1.3
Page 15 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5
Analysis – Overall Performance
This section contains results from analysis of CFRS incident and deployment data using the Incident Analyser Tool (IAT) from a performance perspective. Unless otherwise stated, the graphs relate to deployments occurring in the period 1st January 2003 – 30th August 2006, referred to as the analysis period. The CFRS Emergency Standards are sophisticated. They consider the overall risk for any given station from the perspectives of inherent incident risk and location risk. These are combined to produce bespoke Risk Planes or sets of variable standards for each Station in the form of a Risk Matrix. This report is based on performance against the 2006/07 Risk Planes published in the latest CFRS IRMP Brochure and summarised below.
2006/07 Risk Planes
Station Code 01 02 03 04 05 06 08 09 10 11 12 13 15 16 17 18 19 20 22 23 24 25 26 27 Station WARRINGTON BIRCHWOOD STOCKTON HEATH WIDNES RUNCORN FRODSHAM ELLESMERE PORT CHESTER TARPORLEY MALPAS NANTWICH AUDLEM CREWE SANDBACH HOLMES CHAPEL CONGLETON MACCLESFIELD BOLLINGTON POYNTON WILMSLOW KNUTSFORD NORTHWICH MIDDLEWICH WINSFORD Residential Property Business Property Small Fires SSC (Life Risk) SSC (Non-urgent) Road Traffic Collisions Spillages Leaks UwFDS
Key Very High High Medium Low Very Low Very High High Medium Low Very Low It is certain or very likely that there will be multiple deaths or a single death. It is certain that there will be a major injury, very likely that there will be a single death and likely that there may be multiple deaths. It is certain that there will be a major injury, very likely that there will be an injury and there may be a single death. There will be some disruption with a moderate chance of an injury or loss occurring. There is a reduced chance of disruption with a moderate chance of an injury or loss. Arrive 1 - 5 minutes of turnout on 90% of occasions. Arrive 6 - 10 minutes of turnout on 90% of occasions. Arrive 11 - 15 minutes of turnout on 90% of occasions. Arrive 16 - 20 minutes of turnout on 90% of occasions. Arrive within 21 minutes of turnout on 90% of occasions.
Figure 5.0
Performance against standard at station level in this report is on a ‘Yes’/’No’ basis against the nominated standard. This allows us to aggregate performance at the higher Risk Category, PDG and CFRS levels for comparison purposes.
Version : 1.3
Page 16 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5.1
Response Times
Mobilisation times have in general marginally improved over the analysis period and stand at an average of around 1.9 minutes for CFRS as a whole. Whole Time stations have improved in general over the period and stand at an average of 1.5 minutes. Retained have deteriorated slightly and now stand at an average of 3.8 minutes, but still within target and at a good standard.
Summary Mobilisation Times For All Years
2003
5 4
2004
2005
2006
Time (Minutes
3 2 1 0 Entire Service (RT) Entire Service Entire Service (WT) Entire Service (DM)
Figure 5.1.1 Arrival times have in general deteriorated year on year over the equivalent period and now stand at an average of around 6.7 minutes across CFRS as a whole. Whole Time stations achieve an average of 6.1 minutes over the period with Retained and Day-Crewed achieving averages of 9.5 and 8.3 minutes respectively. This said overall performance against the Cheshire Standards remains very good.
Summary Arrival Times For All Years
2003
11 10 9 8 7 6 5 4 3 2 1 0 Entire Service
2004
2005
2006
Time (Minutes
Entire Service (WT)
Entire Service (RT)
Entire Service (DM)
Figure 5.1.2 There is significant variation across Stations by PDG but this is most likely due to the respective mixes of WT, RT and DC stations in each PDG.
Version : 1.3
Page 17 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5.2
Response Delivery
CFRS performance against their chosen standards is impressive averaging 98% across the Brigade. There appears to have been a marginal year on year decline in performance when compared against the current 2006/07 standards, however previous years performance should really be compared against the standards in force at the time. Retained Stations as a group are currently the best performers overall and they also show a year on year improvement against the 2006/07 standards over the analysis period.
Summary Arrival Delivery For All Years
2003
100.0%
2004
2005
2006
Within Targe
80.0% Entire Service (RT) Entire Service Entire Service (WT) Entire Service (DM)
Figure 5.2.1 Analysis at the PDG level shows consistency across the brigade.
PDG Chester Congleton Crewe and Nantwich Ellesmere Port and Neston Halton Macclesfield Vale Royal Warrington
2003 99.3% 98.9% 96.9% 97.3% 99.1% 97.5% 98.2% 98.6%
2004 98.6% 98.0% 97.1% 96.4% 98.6% 97.7% 97.4% 99.4%
2005 2006 Average 98.3% 98.1% 98.6% 97.9% 100.0% 98.6% 97.5% 95.7% 96.9% 95.7% 95.1% 96.3% 98.8% 98.4% 98.8% 95.7% 96.3% 96.9% 98.2% 96.7% 97.7% 99.4% 98.2% 99.0%
Figure 5.2.2
Overall performance during the study period against the 2006/07 standards at a station level (Figure 5.2.3) are equally impressive, with only a small number of cases of under achievement. In this respect you have to be mindful that where activity levels are low this can have a large impact against the % achievement – these cases require further in-depth analysis
Version : 1.3
Page 18 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
RISK PLANE - All Years Warrington Birchwood Stockton Heath Widnes Runcorn Frodsham Ellesmere Port Chester Tarporley Malpas Nantwich Audlem Crewe Sandbach Holmes Chapel Congleton Macclesfield Bollington Poynton Wilmslow Knutsford Northwich Middlewich Winsford Average
Residential Business Small Fires 97.5% 100.0% 97.1% 98.2% 96.2% 100.0% 93.0% 96.2% 100.0% 100.0% 100.0% 100.0% 89.4% 100.0% 100.0% 95.2% 94.4% 100.0% 97.1% 99.2% 95.8% 97.1% 100.0% 92.6% 96.0% 96.8% 98.2% 100.0% 100.0% 100.0% 100.0% 94.9% 100.0% 100.0% 100.0% 100.0% 100.0% 99.1% 100.0% 92.3% 100.0% 98.6% 100.0% 100.0% 98.7% 100.0% 100.0% 100.0% 94.5% 98.4% 99.0% 99.3% 99.1% 99.0% 99.2% 98.9% 96.4% 99.7% 97.4% 97.0% 97.0% 92.6% 97.5% 96.3% 100.0% 99.6% 98.0% 96.2% 97.8% 98.5% 97.3% 98.9% 97.8% 99.9% 98.5%
SSC - Life Risk 97.5% 100.0% 91.9% 100.0% 89.9% 83.3% 96.4% 98.1% 100.0% 100.0% 100.0% 100.0% 97.3% 100.0% 100.0% 100.0% 96.1% 100.0% 100.0% 92.1% 94.4% 96.4% 100.0% 90.6% 96.2%
SSC - Non Urgent 98.9% 100.0% 100.0% 99.1% 100.0% 90.2% 98.3% 98.9% 92.3% 81.8% 94.8% 88.9% 98.9% 100.0% 100.0% 100.0% 98.1% 96.6% 100.0% 96.1% 94.9% 97.2% 94.1% 96.6% 98.1%
RTC 89.6% 98.2% 90.8% 99.0% 76.6% 94.6% 84.2% 86.8% 84.8% 100.0% 93.2% 95.8% 73.3% 95.6% 92.7% 93.7% 69.6% 100.0% 97.2% 95.3% 51.8% 74.8% 98.1% 83.3% 84.1%
Spillages and Leaks
UFDS
Average 98.8% 99.4% 98.2% 99.1% 98.2% 97.9% 96.2% 98.7% 94.9% 96.7% 98.1% 97.7% 96.1% 98.3% 97.1% 98.4% 96.6% 99.1% 99.1% 98.7% 91.5% 97.3% 98.8% 98.7%
100.0% 99.9% 100.0% 99.7% 100.0% 100.0% 100.0% 99.8% 97.9% 100.0% 100.0% 100.0% 100.0% 99.4% 100.0% 99.9% 100.0% 97.7% 80.0% 95.0% 100.0% 99.7% 100.0% 96.8% 99.8% 100.0% 100.0% 100.0% 98.7% 100.0% 99.3% 100.0% 99.8% 100.0% 100.0% 100.0% 100.0% 100.0% 99.5% 91.7% 99.7% 96.4% 99.6% 100.0% 100.0% 100.0% 100.0% 99.0% 99.8%
Figure 5.2.3 (Blanks represent no data)
Version : 1.3
Page 19 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5.3
Resource Deployment
This section considers resource deployment over the analysis period:
Resource Deployment By Station Ground
2.50
Resources Deploy
2.00
1.50
1.00
0.50
0.00 Middlewich Stockton Heath Tarporley Audlem Poynton Malpas Crewe Macclesfield W arrington Northwich Runcorn Birchwood Congleton Sandbach Bollington Nantwich Frodsham W idnes Knutsford Ellesmere Port Holmes Chapel W ilmslow W insford Chester
Figure 5.3.1 On average 1.6 appliances are deployed from a station to any one particular incident across CFRS, varying from 1.3 for Winsford to 2.2 for Holmes Chapel. The latter statistic can be verified as there is a high level of response to the motorway (M6) which requires a pre-determined response of 3 appliances and therefore provides a reasoned argument for the higher average deployment.
Resource Deployment By Generic Incident
2.50
Resources Deploy
2.00 1.50 1.00 0.50 0.00 162 - Staffordshire Incident 190 - Staffordshire Standby 142 - FDR1 143 - FAM 144 - FAGI 146 - SSC 197 - Derbyshire Incident 194 - GMC Incident 161 - Merseyside Incident 199 - Derbyshire Standby 191 - Merseyside Standby 151 - Shropshire Incident 149 - Chimney Fires 192 - GMC Standby 147 - Small Fires 195 - North W ales Incident 196 - North W ales Standby 198 - Shropshire Standby
Figure 5.3.2 There is considerable variation by generic incident type. The highest deployment of 2.1 appliances (on average) for “144 – False Alarm Good Intent” incidents contrasts with 1.2 for “147 – Small Fires” incidents. 144 Incidents generally relate to premises with AFA systems such as Commercial / Industrial / Care premises, and these require a deployment of 3 appliances on initial call in appropriate cases.
Version : 1.3
Page 20 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5.4
Casualties
This section considers casualties during the analysis period:
Figure 5.4.1 Fire Deaths across CFRS over the analysis period were comparatively low and all civilian. 77% of these deaths were Special Services incidents, and the relatively high number of FDR1 deaths includes those also involving Vehicle fires.
Figure 5.4.2 Non-fatal Brigade injuries (19) represent approximately 0.9% of total injuries over the analysis period.
Version : 1.3
Page 21 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6
Analysis - Scenario Assessment
This section contains results from analysis of CFRS incident and deployment data for specific stations defined in the Modelling Scenarios described in Section 2. Unless otherwise stated, the graphs relate to incidents occurring in the period 1st January 2003 – 30th August 2006, referred to as the analysis period. A report used frequently in this analysis is the ‘Heat Map’ and it is used to show the geographic distribution of incidents by 1km square across the CFRS geographic area. The following ‘Heat Map’ shows all incidents during the analysis period.
Split 23.7%
11.7%
19.2%
13.9%
17.9%
5.5%
5.2%
2.4% 46 500
0.4% 7 750
Count 457
106
268
226
371
345
101
Incidents 1
Version : 1.3
Page 22 of 35
Print Date : 17/11/2006
1500
100
250
10
50
2
5
2
0.1%
Figure 6.0
CFRS : Incident Response Modelling Project
6.1
Birchwood Incident Analysis
The following ‘heat map’ shows the total number of incidents occurring per square kilometre over the four year period in the Birchwood area. The BOLD highlighted squares contain the respective Stations Birchwood (345), Warrington (328) and Stockton Heath (126), which will be used in part within this section as comparators.
15 3 1 4 41 53 16 8 30 90 101 13 1 27 23 7 3 17 24 17 42 14 65 31 23 78 48 14 26 134 102 99 167 55 4 9 8 7 12 12 191 85 233 259 415 85 2 10 15 1 13 1 2 9 73 196 359 431 619 283 80 31 6 7 16 9 213 267 406 328 297 161 75 11 9 26 16 4 11 15 191 459 129 62 236 126 53 14 8 61 1 3 4 68 24 136 127 84 32 154 98 9 9 14 2 87 94 141 172 20 36 16 1 21 92 3 1 24 8 17 134 27 36 117 345 31 145 9 44 34 2 2 20 2 35 9 38 61 14 3 51 12 43 21 28 73 12 2 83 8 3 5 4 7 1 3 10 9 3 7 2
1 5
1 3 1 22 2 12 6 10 23 4 11 61 14 2 8 1 3 35 18 9 3 8
2 1 4 1
9 6 7 19 7 7 1
Figure 6.1.1 (Key as for Figure 6.0)
The profile of Incidents by Hour of Day shows that Birchwood is fairly active between 08:00 and 24:00. Any reconfiguration will have to cater for high demand between 16:00 and 23:00 hours.
Incident Volumes By Station Ground For All Years
Warrington
800 700
Incident Volumes
600 500 400 300 200 100 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
Birchw ood Stockton Heath
Period
Figure 6.1.2
Version : 1.3
Page 23 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
The chart below shows the incident volumes by year for the three stations (2006 is not a complete year). The levels at Birchwood are largely flat (if not showing a marginal increase), whereas the levels at Warrington appear to be reducing overall.
Incident Volumes By Station Ground
Birchw ood
2500
2000
Incident Volumes
Stockton Heath
1500
1000
Warrington
500
0 2003 2004 2005 2006
Period
Figure 6.1.3 The chart below shows the relative proportions (by %) of incident types over this period. Warrington and Birchwood are very similar, with Birchwood experiencing a marginally lower proportion of Small Fires (147) and False Alarm Malicious Intent (143) and marginally higher proportion of FDR1 (142).
Incident Volumes (%) By Station Ground For All Years
100% 90% 80%
147 - Sm all Fires 149 - C him ney Fires
Incident Volum
70% 60% 50% 40% 30% 20% 10%
142 - FD R 1 143 - FAM 144 - FAG I 146 - SSC
0% Birchwood Stockton Heath W arrington
Station Ground
Figure 6.1.4
Version : 1.3
Page 24 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Incident Volumes By Station Ground For All Years
8000 7000
6000 5000 4000 3000 2000 1000 0 Over the Border Macclesfield Knutsford Stockton Heath Tarporley Chester Winsford Runcorn Nantwich Poynton Crewe Northwich Sandbach Congleton Birchwood Middlewich Frodsham Bollington Holmes Chapel Ellesmere Port Warrington Wilmslow Unknown Audlem Widnes Malpas
Warrington Ellesmere Port Chester Runcorn Crew e Widnes Macclesfield Northw ich Winsford Birchw ood Wilmslow Knutsford Stockton Heath Congleton Nantw ich Frodsham Sandbach Tarporley Poynton Middlew ich Bollington Holmes Chapel Malpas Audlem Over the Border Unknow n
Incident Volumes
Station Ground
Figure 6.1.5 The activity level at Birchwood, whilst moderately higher than Stockton Heath, is lower than the two other Day-crewed stations (Winsford and Northwich). This therefore warrants a review of its status which is more akin to that of Day-crewing.
Version : 1.3
Page 25 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6.2
Wilmslow Incident Analysis
The profile of Incidents by Hour of Day show that the bulk of demand on Wilmslow resources is between 08:00 and 24:00 at more than twice the levels experienced outside of these times. The peak loading is between 17:00 and 21:00 and is on average 3 times that of the 00:00 to 08:00 period.
Incident Volumes By Specific Incident For All Years
120
Incident Volumes
100
149 - Chimney Fires
80 60 40 20 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
144 - FAGI 147 - Small Fires
146 - SSC
143 - FAM
142 - FDR1
Period
Figure 6.2.1 It is also interesting to note that primary fires FDR1s (142) are at their highest between 09:00 and 24:00, and are proportionately higher between 18:00 and 24:00, however the overall number of incidents of this nature is relatively low. Small Fires (147) also peak between 12:00 and 24:00 with a proportionately higher share of incidents during this time.
Incident Volumes By Specific Incident For All Years
100%
149 - Chimney Fires
Incident Volumes
80%
147 - Small Fires
60%
146 - SSC
40% 20% 0% 00:00 to 01:00 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
144 - FAGI
143 - FAM
142 - FDR1
Period
Figure 6.2.2 On the face of this initial analysis moving from Wholetime to Day-Crewing/Retained or other response options will expose a volume of incidents to Retained Response times. Therefore further investigation and analysis needs to be undertaken to define the appropriate service delivery mechanism through the incident response simulator.
Version : 1.3
Page 26 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6.3
Northwich / Winsford / Middlewich Incident Analysis
The following ‘heat map’ shows the total number of incidents occurring per square kilometre over the study period in the Northwich / Winsford / Middlewich area. The BOLD highlighted squares contain the respective Stations Northwich (35), Middlewich (89) and Winsford (42).
10 9 6 5 1 92 69 17 28 42 6 1 3 1 2 3 51 3 3 5 2 1 3 3 8 24 16 21 1 5 213 453 2 6 1 4 1 3 59 283 173 193 12 16 56 3 9 2 80 63 3 33 11 6 17 36 72 114 10 2 6 94 237 76 56 15 56 35 49 197 42 1 2 2 1 3 2 78 159 90 35 31 10 7 197 385 32 6 3 116 9 6 3 4 3 5 4 6 10 41 36 135 151 2 6 15 10 28 12 162 1 3 1 3 6 3 24 15 1 2 1 2 147 89 38 2 2 1 18 5 49 27 26 3 2 4 5 4 5 34 55 7 5 5 3 3 6 6 14 5 1 2 5 1 3 5 5 11 1 3 2 5 3 3 2 7 15 20 26 9 10 8 2 1 8 13 1 1 11 44 1 2 5 5 16 25 18 5 2 1 1 43 1 1
Figure 6.3.1 The following charts show total Incident Volume by Station by time of day and Incident Type by time of day respectively. Small Fires (147) account for over 50% of incidents between 16:00 & 22:00. The use of TRV’s may have a significant impact on the mix of appliances required.
Incident Volumes By Station Ground For All Years
250
Incident Volumes
200 150 100 50 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
Middlew ich
Northw ich
Winsford
Period
Figure 6.3.2
Version : 1.3
Page 27 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Incident Volumes By Specific Incident For All Years
450 400 350 300 250 200 150 100 50 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
149 Chimney Fires 147 - Small Fires
Incident Volumes
146 - SSC
144 - FAGI
143 - FAM
142 - FDR1
Period
Figure 6.3.3 It will only be possible to analytically evaluate the impact against Response Standards using the Incident Response Simulator.
Version : 1.3
Page 28 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6.4
Targeted Response Vehicles Analysis
The TRV Response strategy will ultimately be driven by the profile of Small Fire (147) incident demand. It is the response to this demand that needs to be managed effectively to minimise the impact on existing response strategies for other types of incident. Looking at Small Fires (147) by Performance Delivery Group provides a compelling picture of the challenge facing CFRS.
Small Fires (147) by PDG
C hester, 8% W arrington, 21% C ongleton, 3% C rewe and N antwich, 8%
Vale R oyal, 13%
E llesm ere P ort and N eston, 16%
M acclesfield, 6%
H alton, 25%
Figure 6.4.1 Combined incident volumes for Vale Royal, Ellesmere Port & Neston, Warrington and Halton PDGs account for 75% of all Small Fires. This suggests that the greatest return from TRV response options would be obtained by focussing on these four PDG’s. The daily profile of Small Fires (147) incident volumes show demand progressively increasing from 12:00 to a peak at 21:00 before progressively decreasing to 01:00 during a typical day. This represents the outline daily demand profile that a TRV based response option needs to address.
Incident Volumes By Generic Incident For All Years
3000 2500
Incident Volum
2000 1500 1000 500 0 00:00 to 01:00 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
147 - Sm all Fires
Period
Figure 6.4.2
Version : 1.3
Page 29 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6.5
Aerial / Hydraulic Lift Platform Deployment
The following ‘Heat Map’ shows the geographic dispersion of all incidents for the analysis period across CFRS. In particular it highlights the Unitary Authority’s and heavy industrialised areas in the North / North West of the area and the more rural nature of the South / South East.
Figure 6.5.1 (Included for geographic reference)
Figure 6.5.2 (Heat Map of all incidents - Key as Figure 6.0)
Version : 1.3
Page 30 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Comment [PL2]: Is this all incidents and if so should it just be for incidents requiring a HLP appliance?
If the distribution of HP demand is proportionately distributed across CFRS in a similar manner to all incidents then this suggests an initial 3 base solution as:1) 2) 3) One based in the North West conurbations (Chester or Ellesmere Port) One based in the Northern conurbations (Warrington or Halton) One based in the South East (Crewe or Congleton)
Further in depth study of their actual use over the analysis period needs to be undertaken and analysed. Assuming that HP attendance is more closely aligned to FDR1 incidents and knowing that Chester City incidents require a mandatory HP response the following typical daily profile of Primary Fire FDR1 (142) incidents reinforces the above.
Incident Volumes By PDG For All Years
W arrington
Vale R oyal
800 700
M acclesfield
Incident Volum
600
H alton
500 400 300 200 100 0 00:00 to 01:00 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
C hester C rewe and N antwich E llesm ere P ort and N eston
C ongleton
Period
Figure 6.5.3 The overall hourly demand profile also suggests that a 2 base solution between the hours of 04:00 to 12:00 may be possible. Once again the viability of any combination of locations and number of aerials will be evaluated by the Incident Response Simulator against CFRS Emergency Standards.
Version : 1.3
Page 31 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
7
Analysis – Abstraction Data
At present CFRS maintain paper based records for mechanical downtime, however they are in the process of transferring this information into MS Excel based records to assist in analysing this information. At present electronic records exist from the 1st January 2006 to 30th September 2006. Unfortunately appliances going off the run for a few minutes while the mechanic takes a quick look at it may not get notified to Control and hence are not recorded on the sheet.
Data on non-mechanical downtime primarily related to retained appliance availability is more difficult to provide. Currently this information (Stations telephone control to notify the time and date the appliance will be available / not available for the coming day or couple of days) is entered onto the Remsdaq mobilising system and recorded into an 'Events’ file. This records the time and date the appliance is on/off the run. This then automatically updates the required appliances status to available or not available at the required time. This ensures that for mobilisation purposes Control always know the status of that appliance. Unfortunately once the event is actioned by the system it is automatically deleted from the record.
In order to provide statistical data CFRS currently collate retained appliance availability data in an Excel spreadsheet. This records non-availability during the day or night period but does not specify the exact times of the day. It will merely record for example, that between the hours of 09:00-18:00 an appliance was not available for 4 hours – it will not state which 4 hours of that period. The Service will be introducing within the next month or so, a more advanced electronic availability system for the retained stations - this will record for statistical purposes the exact time of the day when the appliance is available / not available.
A brief addendum report will be produced on abstractions if sufficient data is available before the simulation phase is completed. This will look at :Notification of ‘Off the Run’ Failure to Respond Timing of Abstraction Abstractions by day of week
Version : 1.3
Page 32 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
8
Analysis – 8 Riders vs 9 Riders
This section covers a brief analysis undertaken to look at the difference in Incident Times for 2 Pump attendances between 8 and 9 Rider attendances. This analysis is based on 13,478 incidents observed in the period from 1st January 2003 to 30th September 2006. Incident times have been calculated from the “tstn” (Incident start time) and the “tmrav” (Time appliance mobile returning from incident) fields from the CFRS deployment data. The incident time has been defined as the tstn for the first appliance to be notified to the tmrav for the last appliance to leave the incident. 8 or 9 riders have been calculated by combining the Riders for each appliance. Only 8 or 9 rider incidents have been considered. The following charts show the distributions of 8, 9 and combined 8 & 9 Rider incident distributions. The first displays the raw distributions, whilst the second shows a scale-less version based on dividing by the total number of observations in each population.
Raw Distributions
500 450 400 0.0300 350 300 250 200 150 0.0100 100 50 0 1 17 33 49 65 8 Rider 81 97 113 129 145 161 177 193 9 Rider 8 & 9 Rider 0.0050 0.0000 1 17 33 49 65 81 97 113 129 145 161 177 193 8 Rider 9 Rider 8 or 9 Rider 0.0250 0.0200 0.0150 0.0400 0.0350
Scale-less Distributions
Figure 8.1 The following descriptive statistics provide some insight into these incident time distributions and the average incident times for each population:
This information was used to undertake a statistical t-test into the difference between the means of the 8 & 9 rider distributions using a two tailed independent test.
P value and statistical significance: The two-tailed P value equals 0.9536. By conventional criteria, this difference is considered to be not statistically significant. Figure 8.2 The conclusion of this analysis is that based on average incident times alone there is no difference between 8 Rider attended incidents and 9 rider attended incidents.
Comment [PL3]: Can we reduce the number of decimal places in the table please!
Version : 1.3
Page 33 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
9
Conclusions and Next Steps
This section contains the conclusions that we have drawn from the initial analysis and details the next steps.
9.1
Conclusions
1 : High volume of false alarms Around 40% of overall incidents requiring response turn out to be false alarms. It may therefore be possible make a significant impact on overall demand through initiatives designed to reduce the occurrence of false alarms. 2 : No discernible trend in incident numbers There is no discernible trend in overall incident numbers, although last year saw a rise in Special Service Calls. 3 : Seasonality in incident volumes There is seasonality in incident volumes. The simulator will be able to show whether there is either a need for greater resources in busier months or spare capacity that can safely be reduced in quieter months through expansion of the current strategic reserve. Alternatively, the model may show that it is not possible to reduce capacity in the quieter months due to the need to maintain coverage at all times. 4 : Significant variance in overall incident workload (Small Fires Peak) There is a significant variance in overall incident workload through the day. The Total volume of Incidents per hour is flat between 09:00 and 23:00, however Small Fires (147) can more than double this volume between 17:00 and 22:00 hours. The active use of TRV’s during these periods should be evaluated and encouraged. 5 : Emergency Response Performance CFRS Emergency Response performance against the 2006/07 Emergency Response standards is impressive. At a Brigade level an average of 98% attendance within standard has been achieved over the study period. 6 : 8 vs 9 Rider attendances The conclusion of this analysis is that based on average incident times alone there is no discernable difference between 8 Rider attended incidents and 9 rider attended incidents.
9.2
o o o
Next Steps.
Configure the Incident Response Simulator to use CFRS Incident Data. Calibrate the Simulator against known CFRS performance. Design an experimentation plan to evaluate the five scenarios identified above.
The next steps are as follows:
Version : 1.3
Page 34 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Appendix 1 – Glossary of Terms
This section covers basic definitions and terminology relevant to this report.
Fire Incident Types
Incident Code 142 143 144 146 147 149 151 161 162 190 191 192 194 195 196 197 198 199 Type Cheshire Incident Cheshire Incident Cheshire Incident Cheshire Incident Cheshire Incident Cheshire Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Description FDR1 (Structural Property Fires) FAM (False Alarm Malicious) FAGI (False Alarm Good Intent) SSC (Special Service Calls) Small Fires Chimney Fires Shropshire Incident Merseyside Incident Staffordshire Incident Staffordshire Standby Merseyside Standby GMC Standby GMC Incident Northwales Incident Northwales Standby Derbyshire Incident Shropshire Standby Derbyshire Standby
Version : 1.3
Page 35 of 35
Print Date : 17/11/2006
doc_319101559.pdf
Incident management (IcM) is a term describing the activities of an organization to identify, analyze, and correct hazards to prevent a future re-occurrence. These incidents within a structured organization are normally dealt with by either an Incident Response Team (IRT), or an Incident Management Team (IMT).
CFRS : Incident Response Modelling Project
Incident Response Modelling Project
Cheshire Fire and Rescue Service (CFRS).
Richard Smart Process Evolution Limited Version 1.3 17th November 2006
Tel: 07782 170337 Fax: 08701 709887 E-mail: [email protected] Web: www.processevolution.co.uk Process Evolution Limited, 6 Fordbank Court, 218 Henwick Road, Worcester. WR2 5PF Registered in England No. 4983030, VAT 823-9689-84
CFRS : Incident Response Modelling Project
Executive Summary
This document forms the Interim Report for Cheshire Fire & Rescue Service (CFRS) as part of the Incident Response Modelling Project being conducted by Process Evolution on their behalf. This report is a key deliverable for the “Diagnose” phase of the project. It contains the results of initial analysis of data captured from CFRS operational systems and has two main purposes: 1) 2) To ensure an analytical understanding of CFRS operations and data prior to commencing the “Evaluate” phase of the project. To provide CFRS with detailed analytical insight into a range of proposed modelling scenarios prior to evaluation.
The report provides a high level analysis of Incident Volumes and Operational Performance of CFRS based on Incident and Deployment data between 1st January 2003 and 30th September 2006. It also provides initial insight into 5 CFRS specified modelling scenarios which will be the subject of further simulation assessment in subsequent stages of this project: Birchwood Fire Station Wilmslow Fire Station Northwich / Winsford / Middlewich Fire Station Reconfiguration Targetted Response Vehicles (TRVs) Deployment Aerial / Hydraulic Lift Platform Deployment
Summary conclusions to be drawn from this analysis are: 1 High volume of incidents requiring attendance are false alarms (over 40%). The Incident Response Simulator could be used to look at the effects of investigative attendance strategies on resource loading. No clear trend in overall incident numbers, although they appear to be reducing. There is clear seasonality in incident volumes. The Incident Response Simulator could be used to investigate a range of appliance rostering solutions. Significant variance in overall daily incident workload (Small Fires in particular). This suggests greater use could be made of TRV’s during peak periods. Emergency Response Performance CFRS are achieving a 98% (3½ year average) Emergency Response performance against the 2006/07 Emergency Response standards at Brigade level. No discernable difference between 8 vs 9 Rider attendances on Incident Times.
2 3 4 5
6
These conclusions are discussed in greater detail in the body of the report. The next steps are to complete the ‘Evaluate’ Phase as follows: - Configure the Incident Response Simulator to use CFRS Incident Data. - Calibrate the Simulator against known CFRS performance. - Design an experimentation plan to evaluate the five scenarios identified above.
Version : 1.3
Page 2 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Table of Contents
Executive Summary .......................................................................................................... 2 Table of Contents ................................................................................................................ 3 Document History ................................................................................................................ 3 1 2 3 4
4.1 4.2 4.3 4.4 4.5 4.6
Introduction................................................................................................................. 4 Background................................................................................................................. 5 Approach..................................................................................................................... 7 Analysis – Overall Incident Volumes ........................................................................ 8
Incidents by Type.................................................................................................................................... 8 Incidents by Year .................................................................................................................................. 10 Incidents by Month................................................................................................................................ 12 Incidents by Day of Week ..................................................................................................................... 13 Incidents by Hour of Day....................................................................................................................... 14 Incidents by Station .............................................................................................................................. 15
5
5.1 5.2 5.3 5.4
Analysis – Overall Performance .............................................................................. 16
Response Times................................................................................................................................... 17 Response Delivery................................................................................................................................ 18 Resource Deployment .......................................................................................................................... 20 Casualties............................................................................................................................................. 21
6 7 8 9
Analysis - Scenario Assessment............................................................................. 22 Analysis – Abstraction Data .................................................................................... 32 Analysis – 8 Riders vs 9 Riders............................................................................... 33 Conclusions and Next Steps ................................................................................... 34
9.1 Conclusions .......................................................................................................................................... 34 9.2 Next Steps. ........................................................................................................................................... 34
Appendix 1 – Glossary of Terms.................................................................................... 35
Document History
Version
0.1 1.0 1.1 1.2 1.3
Release
Initial Draft Interim Report Final Release Candidate (Post completion of IAT & Risk Planes) Final Release Candidate Final Release Final Release
Date
10 October 2006 2nd November 2006 5th November 2006 14th November 2006 17th November 2006
th
Version : 1.3
Page 3 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
1
Introduction
This document forms the Interim Report for Cheshire Fire & Rescue Service (CFRS) as part of the Incident Response Modelling Project being conducted by Process Evolution on their behalf. This report is a key deliverable for the “Diagnose” phase of the project. It contains the results of initial analysis of data captured from CFRS’s operational systems and has two main purposes: 1) 2) To ensure an analytical understanding of CFRS operations and data prior to commencing the “Evaluate” phase of the project. To provide CFRS with detailed analytical insight into a range of proposed modelling scenarios prior to evaluation.
Version : 1.3
Page 4 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
2
Background
The Government’s National Framework Strategy challenges fire services to review the deployment of their resources in order to provide targeted response in a more cost effective manner. Previous incident response standards and corresponding deployment strategies are based on historical guidelines and risks. As part of its response to these challenges CFRS is considering a number of changes to its deployment strategy for emergency response, such as the: • • • Mix between whole time (WT), retained (RT) and day-crewed (DC) appliances. Identifying optimal station bases for specialist appliances. Crewing/use of appropriate response vehicles such as Targeted Response Vehicles.
Process Evolution specialises in the design, analysis and continuous improvement of business processes. A key feature of our work is the use of quantitative techniques which enable the impact of proposed changes to be quantified prior to implementation. CFRS have therefore commissioned Process Evolution to undertake initial evaluations of specific proposed changes, and to help the service establish an ‘in house’ capability to evaluate such changes in the future. The specific proposed changes to be evaluated are: Birchwood Fire Station Quantify the impact of providing cover to the Birchwood area from Warrington at certain times of day or changing availability from Whole Time to Day-crewed, or other appropriate response options. Wilmslow Fire Station Evaluate the effect of changing availability from Whole Time to Day Crewing / Retained, or other appropriate response options. Northwich / Winsford / Middlewich Fire Station Reconfiguration Evaluate the effects of changing appliance availability mix across these stations. In particular, it may be possible to provide a reduced level of immediately available cover at certain times, with a 1 hour response say for additional appliances in the event of a major incident, or other appropriate response options. Targetted Response Vehicles (TRVs) Deployment Identify the station bases and operating model for TRV’s and corresponding hours worked that will deliver optimal response times across CFRS. Aerial / Hydraulic Lift Platform Deployment Identify the station bases for HP’s that will deliver optimal response times across CFRS. Look at effect of reducing availability of one HP to retained.
For performance purposes CFRS is structured into 8 Performance Delivery Groups (PDGs) one for each of its Local Government Districts and Unitary Authorities. These are individually colour coded on the map on the following page.
Version : 1.3
Page 5 of 35
Print Date : 17/11/2006
Version : 1.3
Performance Delivery Groups
Chester Congleton Crewe and Nantwich Ellesmere Port and Neston Halton Macclesfield Vale Royal Warrington
CFRS Station Map
CFRS : Incident Response Modelling Project
Page 6 of 35
PDG
Figure 2.1
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
3
Approach
The diagram below shows the main steps in the project.
Consulting Approach
Monitor performance
Agree Proposal
Agree Project Scope
Diagnose
Evaluate
Plan
Handover
Provide Support
Deliverables Terms of Reference Interim Report Model Runs Report Final Report
Analysis Toolkit
Figure 3.1
The “Scoping” phase resulted in the identification of the five scenarios detailed for initial evaluation and summarised in the previous section. This document reports on the ‘Diagnose’ phase of the project. During this phase, we extracted information from CFRS’s command and control system relating to incidents occurring in the calendar years 2003 – 2006, providing 3 ½ years of historical information referred to as the analysis period throughout this document. This forms a statistically robust and representative sample of data on which to undertake the analysis for this project. We have configured our Incident Analysis Toolkit (IAT) to import this data and have used it to provide the analysis contained in Section 4. Additional discussions have been held with CFRS staff in order to further understand the deployment process. This has been incorporated in the Simulator. Work has already begun on configuring our Incident Response Simulator to reflect the infrastructure of CFRS in terms of the number and location of fire stations and the respective appliances based at each one. During the ‘Evaluate’ phase, we shall use this model to test the proposed changes and quantify the impact on response performance against CFRS’s response standards for first attendance at incidents. These will be based on the 2006/07 Response Standards published by CFRS for each Risk Plane associated with a Station.
Version : 1.3
Page 7 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4
Analysis – Overall Incident Volumes
This section contains results from analysis of CFRS incident data using the Incident Analyser Tool (IAT). Unless otherwise stated, the graphs relate to incidents occurring in the period 1st January 2003 – 30th August 2006 which is referred to as the analysis period.
4.1
Incidents by Type
The chart shows the total number of incidents of each specific type by PDG Area that occurred during the analysis period:
Incident Volumes By Specific Incident For All Years
25000
W arrington
Vale R oyal
20000
M acclesfield
Incident Volum
H alton
15000
E llesm ere P ort and N eston
10000
C rewe and N antwich
5000
C ongleton
0 146 - SSC 142 - FDR1 144 - FAGI 143 - FAM 149 - Chimney Fires 147 - Small Fires
C hester
Specific Incident
Figure 4.1.1 It can be seen that the largest category is Small Fires (147) closely followed by False Alarms Good Intent (144).
Incident Volumes By PDG For All Years
14000 12000
149 - C him ney Fires
147 - Sm all Fires
Incident Volum
10000 8000 6000 4000 2000 0 Macclesfield Halton Ellesmere Port and Neston Crewe and Nantwich Warrington Vale Royal Congleton Chester
143 - FAM 146 - SSC
144 - FAG I
142 - FD R 1
PDG
Figure 4.1.2 Volumes by PDG vary considerably with Warrington and Halton being almost 60% higher than other districts. Types of incident (by %) also vary considerably by PDG as can be seen in the following chart.
Version : 1.3
Page 8 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Incident Volumes By PDG For All Years
100% 90% 80%
149 - C him ney Fires
Incident Volum
147 - Sm all Fires
70% 60% 50% 40% 30% 20% 10% 0% Macclesfield Halton Ellesmere Port and Neston Crewe and Nantwich W arrington Vale Royal Congleton Chester
142 - FD R 1 143 - FAM 144 - FAGI 146 - SSC
PDG
Figure 4.1.3 Small Fires pre-occupy the urban areas of Halton, Warrington, Ellesmere Port and Vale Royal, where as the more rural areas deal with more False Alarm Good Intents (144) and Special Services (146). The greater proportion of “Over the Border” incidents tend to be Special Services (146)
Version : 1.3
Page 9 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.2
Incidents by Year
The next chart shows the number of incidents by PDG over the analysis period:
Incident Volumes By PDG
20000
W arrington
18000 16000
Vale R oyal
Incident Volum
14000 12000 10000
M acclesfield
H alton
E llesm ere P ort and N eston
8000 6000 4000 2000
C hester C rewe and N antwich
C ongleton
0 2003 2004 2005 2006
Period
Figure 4.2.1 There is no clear trend in incident volumes over time however there is a suggestion from the 2003/04/05 that volumes are decreasing. This cannot be determined until a complete set of incident statistics are available for 2006.
Incident Volumes By PDG
4000 3500 3000
W arrington V ale Royal M acclesfield
Incident Volum
2500
Halton
2000
Ellesmere Port and Neston
1500
Crewe and Nantwich
1000 500 0 2003 2004 2005 2006
Congleton Chester
Period
Figure 4.2.2
Version : 1.3
Page 10 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
If we examine this data by incident type, we can see from the chart below (Figure 4.2.3) that most of the decrease from 2003 to 2004/05 is accounted for by a drop in Small Fires (147).
Incident Volumes By Specific Incident
149 - Chimney Fires
20000 18000 16000
147 - Small Fires
Incident Volumes
14000 12000 10000 8000 6000 4000 2000 0 2003 2004 2005 2006
142 - FDR1 143 - FAM 144 - FAGI 146 - SSC
Period
Figure 4.2.3
Version : 1.3
Page 11 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.3
Incidents by Month
The next chart suggests there is some seasonality in the number of incidents – however with an incomplete 2006 data set this is difficult to conclude. The busiest months are July and August, with Small Fires (147), SSC’s (146) and False Alarm Good Intents (144) being the main contributors to these rises in activity levels..
Incident Volumes By Specific Incident For All Years
7000 6000
149 - Chimney Fires
147 - Small Fires
Comment [PL1]: Suggest that the effect of incomplete 2006 is removed by pro-rata or eliminating 2006 data – do this throughout
146 - SSC
Incident Volumes
5000
144 - FAGI
4000 3000 2000
142 - FDR1 143 - FAM
1000 0 September November June January March July October December April May February August
Period
Figure 4.3.1 Incident types vary considerably by month. Chimney Fires (149) only show during the winter months (Nov-Apr) and False Alarms Malicious Intent (143) appear to be proportionately higher during those months. Incident volumes appear to be around 40% higher during the summer months, though there is some volatility around holiday periods.
Incident Volumes By Specific Incident For All Years
100%
149 - Chimney Fires
147 - Small Fires
80%
146 - SSC
Incident Volumes
60%
144 - FAGI
40%
143 - FAM
20%
142 - FDR1
0% September November June January July October December March April February August May
Period
Figure 4.3.2 It is no surprise that Chimney fires occur primarily in the winter months (Nov-Apr) December to March whereas Small Fires (147) tend to be more frequent in the summer months.
Version : 1.3
Page 12 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.4
Incidents by Day of Week
Weekends appear marginally busier than other days of the week:
Incident Volumes By PDG For All Years
10000 9000 8000
W arrington
Vale R oyal
Incident Volum
7000 6000 5000 4000 3000 2000 1000 0 Thursday Tuesday Friday Saturday Monday W ednesday Sunday
M acclesfield
H alton
E llesm ere P ort and N eston
C rewe and N antwich
C ongleton
C hester
Period
Figure 4.4.1 However, volumes on the busiest day (Saturday) are less than 10% more than on the quietest day, which does not in itself suggest the need for a widely different deployment capability on different days of the week. Looking at the proportions of Incident Type by day reveals a subtly different pattern of behaviour. rds There appears to be a 1/3 reduction in the proportion of False Alarms Good Intent (144) at weekends, whereas there is a 20% increase in Small Fires (147) over the same period.
Incident Volumes By Specific Incident For All Years
149 - C him ney Fires
100% 90%
147 - Sm all Fires
80%
Incident Volum
70% 60% 50% 40% 30% 20% 10% 0% Tuesday Thursday Friday Saturday Monday W ednesday Sunday
142 - FD R 1 143 - FAM 144 - FAG I 146 - SSC
Period
Figure 4.4.2 This suggests a need to hold greater Small Fire capacity for weekend cover.
Version : 1.3
Page 13 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.5
Incidents by Hour of Day
The next chart shows when incidents occur through the day over the analysis period:
Incident Volumes By Generic Incident For All Years
5000 4500 4000
149 - C him ney Fires
Incident Volum
3500 3000 2500 2000 1500 1000 500 0 00:00 to 01:00 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
147 - Sm all Fires
146 - SSC
144 - FAG I
143 - FAM
142 - FD R 1
Period
Figure 4.5.1 Here the variance over time is much more striking, with peak demand between 17:00 and 21:00 over six times that between 05:00-06:00. It can also be seen that different incident types have different peak periods. False Alarms Good rd Intent (144) primarily occur from 08:00-22:00 and make up over 2/3 s of incidents between 09:00-10:00, whereas False Alarms Malicious Intent (143) peak between 18:00 and 23:00. Small Fires grow progressively between 12:00 and 24:00 peaking around 19:00-21:00 and account for over 50% of calls at that time of day.
Incident Volumes By Specific Incident For All Years
600 500
Incident Volumes
400 300 200 100 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
142 - FDR1
Period
Figure 4.5.2 This chart shows that Primary Fires FDR1 (142) peak between 20:00 and 23:00 at around 3½ times that of the lowest demand between 05:00-08:00.
This variation in incident demand suggests that some benefit may be derived from looking at appliance shift patterns and matching them more closely to the daily demand profiles illustrated in these two charts.
Version : 1.3
Page 14 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
4.6
Incidents by Station
The next chart shows the proportion of incidents that occur at each station, together with colour coding to indicate the crewing status of each station. Whole-time Stations cover the highest volume of incidents, though Birchwood and Wilmslow experience lower demand than the Day-crewed Northwich and Winsford Stations.
Incident Volumes By Station Ground For All Years
8000 7000
6000 5000 4000 3000 2000 1000 0 Over the Border Macclesfield Knutsford Stockton Heath Tarporley Chester Winsford Northwich Sandbach Nantwich Poynton Runcorn Crewe Congleton Birchwood Middlewich Bollington Holmes Chapel Frodsham Ellesmere Port Warrington Wilmslow Unknown Audlem Widnes Malpas
Warrington Ellesmere Port Chester Runcorn Crew e Widnes Macclesfield Northw ich Winsford Birchw ood Wilmslow Knutsford Stockton Heath Congleton Nantw ich Frodsham Sandbach Tarporley Poynton Middlew ich Bollington Holmes Chapel Malpas Audlem Over the Border Unknow n
Incident Volumes
Station Ground
Figure 4.6.1
Station Key Whole-time Day-crewed Retained
Version : 1.3
Page 15 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5
Analysis – Overall Performance
This section contains results from analysis of CFRS incident and deployment data using the Incident Analyser Tool (IAT) from a performance perspective. Unless otherwise stated, the graphs relate to deployments occurring in the period 1st January 2003 – 30th August 2006, referred to as the analysis period. The CFRS Emergency Standards are sophisticated. They consider the overall risk for any given station from the perspectives of inherent incident risk and location risk. These are combined to produce bespoke Risk Planes or sets of variable standards for each Station in the form of a Risk Matrix. This report is based on performance against the 2006/07 Risk Planes published in the latest CFRS IRMP Brochure and summarised below.
2006/07 Risk Planes
Station Code 01 02 03 04 05 06 08 09 10 11 12 13 15 16 17 18 19 20 22 23 24 25 26 27 Station WARRINGTON BIRCHWOOD STOCKTON HEATH WIDNES RUNCORN FRODSHAM ELLESMERE PORT CHESTER TARPORLEY MALPAS NANTWICH AUDLEM CREWE SANDBACH HOLMES CHAPEL CONGLETON MACCLESFIELD BOLLINGTON POYNTON WILMSLOW KNUTSFORD NORTHWICH MIDDLEWICH WINSFORD Residential Property Business Property Small Fires SSC (Life Risk) SSC (Non-urgent) Road Traffic Collisions Spillages Leaks UwFDS
Key Very High High Medium Low Very Low Very High High Medium Low Very Low It is certain or very likely that there will be multiple deaths or a single death. It is certain that there will be a major injury, very likely that there will be a single death and likely that there may be multiple deaths. It is certain that there will be a major injury, very likely that there will be an injury and there may be a single death. There will be some disruption with a moderate chance of an injury or loss occurring. There is a reduced chance of disruption with a moderate chance of an injury or loss. Arrive 1 - 5 minutes of turnout on 90% of occasions. Arrive 6 - 10 minutes of turnout on 90% of occasions. Arrive 11 - 15 minutes of turnout on 90% of occasions. Arrive 16 - 20 minutes of turnout on 90% of occasions. Arrive within 21 minutes of turnout on 90% of occasions.
Figure 5.0
Performance against standard at station level in this report is on a ‘Yes’/’No’ basis against the nominated standard. This allows us to aggregate performance at the higher Risk Category, PDG and CFRS levels for comparison purposes.
Version : 1.3
Page 16 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5.1
Response Times
Mobilisation times have in general marginally improved over the analysis period and stand at an average of around 1.9 minutes for CFRS as a whole. Whole Time stations have improved in general over the period and stand at an average of 1.5 minutes. Retained have deteriorated slightly and now stand at an average of 3.8 minutes, but still within target and at a good standard.
Summary Mobilisation Times For All Years
2003
5 4
2004
2005
2006
Time (Minutes
3 2 1 0 Entire Service (RT) Entire Service Entire Service (WT) Entire Service (DM)
Figure 5.1.1 Arrival times have in general deteriorated year on year over the equivalent period and now stand at an average of around 6.7 minutes across CFRS as a whole. Whole Time stations achieve an average of 6.1 minutes over the period with Retained and Day-Crewed achieving averages of 9.5 and 8.3 minutes respectively. This said overall performance against the Cheshire Standards remains very good.
Summary Arrival Times For All Years
2003
11 10 9 8 7 6 5 4 3 2 1 0 Entire Service
2004
2005
2006
Time (Minutes
Entire Service (WT)
Entire Service (RT)
Entire Service (DM)
Figure 5.1.2 There is significant variation across Stations by PDG but this is most likely due to the respective mixes of WT, RT and DC stations in each PDG.
Version : 1.3
Page 17 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5.2
Response Delivery
CFRS performance against their chosen standards is impressive averaging 98% across the Brigade. There appears to have been a marginal year on year decline in performance when compared against the current 2006/07 standards, however previous years performance should really be compared against the standards in force at the time. Retained Stations as a group are currently the best performers overall and they also show a year on year improvement against the 2006/07 standards over the analysis period.
Summary Arrival Delivery For All Years
2003
100.0%
2004
2005
2006
Within Targe
80.0% Entire Service (RT) Entire Service Entire Service (WT) Entire Service (DM)
Figure 5.2.1 Analysis at the PDG level shows consistency across the brigade.
PDG Chester Congleton Crewe and Nantwich Ellesmere Port and Neston Halton Macclesfield Vale Royal Warrington
2003 99.3% 98.9% 96.9% 97.3% 99.1% 97.5% 98.2% 98.6%
2004 98.6% 98.0% 97.1% 96.4% 98.6% 97.7% 97.4% 99.4%
2005 2006 Average 98.3% 98.1% 98.6% 97.9% 100.0% 98.6% 97.5% 95.7% 96.9% 95.7% 95.1% 96.3% 98.8% 98.4% 98.8% 95.7% 96.3% 96.9% 98.2% 96.7% 97.7% 99.4% 98.2% 99.0%
Figure 5.2.2
Overall performance during the study period against the 2006/07 standards at a station level (Figure 5.2.3) are equally impressive, with only a small number of cases of under achievement. In this respect you have to be mindful that where activity levels are low this can have a large impact against the % achievement – these cases require further in-depth analysis
Version : 1.3
Page 18 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
RISK PLANE - All Years Warrington Birchwood Stockton Heath Widnes Runcorn Frodsham Ellesmere Port Chester Tarporley Malpas Nantwich Audlem Crewe Sandbach Holmes Chapel Congleton Macclesfield Bollington Poynton Wilmslow Knutsford Northwich Middlewich Winsford Average
Residential Business Small Fires 97.5% 100.0% 97.1% 98.2% 96.2% 100.0% 93.0% 96.2% 100.0% 100.0% 100.0% 100.0% 89.4% 100.0% 100.0% 95.2% 94.4% 100.0% 97.1% 99.2% 95.8% 97.1% 100.0% 92.6% 96.0% 96.8% 98.2% 100.0% 100.0% 100.0% 100.0% 94.9% 100.0% 100.0% 100.0% 100.0% 100.0% 99.1% 100.0% 92.3% 100.0% 98.6% 100.0% 100.0% 98.7% 100.0% 100.0% 100.0% 94.5% 98.4% 99.0% 99.3% 99.1% 99.0% 99.2% 98.9% 96.4% 99.7% 97.4% 97.0% 97.0% 92.6% 97.5% 96.3% 100.0% 99.6% 98.0% 96.2% 97.8% 98.5% 97.3% 98.9% 97.8% 99.9% 98.5%
SSC - Life Risk 97.5% 100.0% 91.9% 100.0% 89.9% 83.3% 96.4% 98.1% 100.0% 100.0% 100.0% 100.0% 97.3% 100.0% 100.0% 100.0% 96.1% 100.0% 100.0% 92.1% 94.4% 96.4% 100.0% 90.6% 96.2%
SSC - Non Urgent 98.9% 100.0% 100.0% 99.1% 100.0% 90.2% 98.3% 98.9% 92.3% 81.8% 94.8% 88.9% 98.9% 100.0% 100.0% 100.0% 98.1% 96.6% 100.0% 96.1% 94.9% 97.2% 94.1% 96.6% 98.1%
RTC 89.6% 98.2% 90.8% 99.0% 76.6% 94.6% 84.2% 86.8% 84.8% 100.0% 93.2% 95.8% 73.3% 95.6% 92.7% 93.7% 69.6% 100.0% 97.2% 95.3% 51.8% 74.8% 98.1% 83.3% 84.1%
Spillages and Leaks
UFDS
Average 98.8% 99.4% 98.2% 99.1% 98.2% 97.9% 96.2% 98.7% 94.9% 96.7% 98.1% 97.7% 96.1% 98.3% 97.1% 98.4% 96.6% 99.1% 99.1% 98.7% 91.5% 97.3% 98.8% 98.7%
100.0% 99.9% 100.0% 99.7% 100.0% 100.0% 100.0% 99.8% 97.9% 100.0% 100.0% 100.0% 100.0% 99.4% 100.0% 99.9% 100.0% 97.7% 80.0% 95.0% 100.0% 99.7% 100.0% 96.8% 99.8% 100.0% 100.0% 100.0% 98.7% 100.0% 99.3% 100.0% 99.8% 100.0% 100.0% 100.0% 100.0% 100.0% 99.5% 91.7% 99.7% 96.4% 99.6% 100.0% 100.0% 100.0% 100.0% 99.0% 99.8%
Figure 5.2.3 (Blanks represent no data)
Version : 1.3
Page 19 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5.3
Resource Deployment
This section considers resource deployment over the analysis period:
Resource Deployment By Station Ground
2.50
Resources Deploy
2.00
1.50
1.00
0.50
0.00 Middlewich Stockton Heath Tarporley Audlem Poynton Malpas Crewe Macclesfield W arrington Northwich Runcorn Birchwood Congleton Sandbach Bollington Nantwich Frodsham W idnes Knutsford Ellesmere Port Holmes Chapel W ilmslow W insford Chester
Figure 5.3.1 On average 1.6 appliances are deployed from a station to any one particular incident across CFRS, varying from 1.3 for Winsford to 2.2 for Holmes Chapel. The latter statistic can be verified as there is a high level of response to the motorway (M6) which requires a pre-determined response of 3 appliances and therefore provides a reasoned argument for the higher average deployment.
Resource Deployment By Generic Incident
2.50
Resources Deploy
2.00 1.50 1.00 0.50 0.00 162 - Staffordshire Incident 190 - Staffordshire Standby 142 - FDR1 143 - FAM 144 - FAGI 146 - SSC 197 - Derbyshire Incident 194 - GMC Incident 161 - Merseyside Incident 199 - Derbyshire Standby 191 - Merseyside Standby 151 - Shropshire Incident 149 - Chimney Fires 192 - GMC Standby 147 - Small Fires 195 - North W ales Incident 196 - North W ales Standby 198 - Shropshire Standby
Figure 5.3.2 There is considerable variation by generic incident type. The highest deployment of 2.1 appliances (on average) for “144 – False Alarm Good Intent” incidents contrasts with 1.2 for “147 – Small Fires” incidents. 144 Incidents generally relate to premises with AFA systems such as Commercial / Industrial / Care premises, and these require a deployment of 3 appliances on initial call in appropriate cases.
Version : 1.3
Page 20 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
5.4
Casualties
This section considers casualties during the analysis period:
Figure 5.4.1 Fire Deaths across CFRS over the analysis period were comparatively low and all civilian. 77% of these deaths were Special Services incidents, and the relatively high number of FDR1 deaths includes those also involving Vehicle fires.
Figure 5.4.2 Non-fatal Brigade injuries (19) represent approximately 0.9% of total injuries over the analysis period.
Version : 1.3
Page 21 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6
Analysis - Scenario Assessment
This section contains results from analysis of CFRS incident and deployment data for specific stations defined in the Modelling Scenarios described in Section 2. Unless otherwise stated, the graphs relate to incidents occurring in the period 1st January 2003 – 30th August 2006, referred to as the analysis period. A report used frequently in this analysis is the ‘Heat Map’ and it is used to show the geographic distribution of incidents by 1km square across the CFRS geographic area. The following ‘Heat Map’ shows all incidents during the analysis period.
Split 23.7%
11.7%
19.2%
13.9%
17.9%
5.5%
5.2%
2.4% 46 500
0.4% 7 750
Count 457
106
268
226
371
345
101
Incidents 1
Version : 1.3
Page 22 of 35
Print Date : 17/11/2006
1500
100
250
10
50
2
5
2
0.1%
Figure 6.0
CFRS : Incident Response Modelling Project
6.1
Birchwood Incident Analysis
The following ‘heat map’ shows the total number of incidents occurring per square kilometre over the four year period in the Birchwood area. The BOLD highlighted squares contain the respective Stations Birchwood (345), Warrington (328) and Stockton Heath (126), which will be used in part within this section as comparators.
15 3 1 4 41 53 16 8 30 90 101 13 1 27 23 7 3 17 24 17 42 14 65 31 23 78 48 14 26 134 102 99 167 55 4 9 8 7 12 12 191 85 233 259 415 85 2 10 15 1 13 1 2 9 73 196 359 431 619 283 80 31 6 7 16 9 213 267 406 328 297 161 75 11 9 26 16 4 11 15 191 459 129 62 236 126 53 14 8 61 1 3 4 68 24 136 127 84 32 154 98 9 9 14 2 87 94 141 172 20 36 16 1 21 92 3 1 24 8 17 134 27 36 117 345 31 145 9 44 34 2 2 20 2 35 9 38 61 14 3 51 12 43 21 28 73 12 2 83 8 3 5 4 7 1 3 10 9 3 7 2
1 5
1 3 1 22 2 12 6 10 23 4 11 61 14 2 8 1 3 35 18 9 3 8
2 1 4 1
9 6 7 19 7 7 1
Figure 6.1.1 (Key as for Figure 6.0)
The profile of Incidents by Hour of Day shows that Birchwood is fairly active between 08:00 and 24:00. Any reconfiguration will have to cater for high demand between 16:00 and 23:00 hours.
Incident Volumes By Station Ground For All Years
Warrington
800 700
Incident Volumes
600 500 400 300 200 100 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
Birchw ood Stockton Heath
Period
Figure 6.1.2
Version : 1.3
Page 23 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
The chart below shows the incident volumes by year for the three stations (2006 is not a complete year). The levels at Birchwood are largely flat (if not showing a marginal increase), whereas the levels at Warrington appear to be reducing overall.
Incident Volumes By Station Ground
Birchw ood
2500
2000
Incident Volumes
Stockton Heath
1500
1000
Warrington
500
0 2003 2004 2005 2006
Period
Figure 6.1.3 The chart below shows the relative proportions (by %) of incident types over this period. Warrington and Birchwood are very similar, with Birchwood experiencing a marginally lower proportion of Small Fires (147) and False Alarm Malicious Intent (143) and marginally higher proportion of FDR1 (142).
Incident Volumes (%) By Station Ground For All Years
100% 90% 80%
147 - Sm all Fires 149 - C him ney Fires
Incident Volum
70% 60% 50% 40% 30% 20% 10%
142 - FD R 1 143 - FAM 144 - FAG I 146 - SSC
0% Birchwood Stockton Heath W arrington
Station Ground
Figure 6.1.4
Version : 1.3
Page 24 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Incident Volumes By Station Ground For All Years
8000 7000
6000 5000 4000 3000 2000 1000 0 Over the Border Macclesfield Knutsford Stockton Heath Tarporley Chester Winsford Runcorn Nantwich Poynton Crewe Northwich Sandbach Congleton Birchwood Middlewich Frodsham Bollington Holmes Chapel Ellesmere Port Warrington Wilmslow Unknown Audlem Widnes Malpas
Warrington Ellesmere Port Chester Runcorn Crew e Widnes Macclesfield Northw ich Winsford Birchw ood Wilmslow Knutsford Stockton Heath Congleton Nantw ich Frodsham Sandbach Tarporley Poynton Middlew ich Bollington Holmes Chapel Malpas Audlem Over the Border Unknow n
Incident Volumes
Station Ground
Figure 6.1.5 The activity level at Birchwood, whilst moderately higher than Stockton Heath, is lower than the two other Day-crewed stations (Winsford and Northwich). This therefore warrants a review of its status which is more akin to that of Day-crewing.
Version : 1.3
Page 25 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6.2
Wilmslow Incident Analysis
The profile of Incidents by Hour of Day show that the bulk of demand on Wilmslow resources is between 08:00 and 24:00 at more than twice the levels experienced outside of these times. The peak loading is between 17:00 and 21:00 and is on average 3 times that of the 00:00 to 08:00 period.
Incident Volumes By Specific Incident For All Years
120
Incident Volumes
100
149 - Chimney Fires
80 60 40 20 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
144 - FAGI 147 - Small Fires
146 - SSC
143 - FAM
142 - FDR1
Period
Figure 6.2.1 It is also interesting to note that primary fires FDR1s (142) are at their highest between 09:00 and 24:00, and are proportionately higher between 18:00 and 24:00, however the overall number of incidents of this nature is relatively low. Small Fires (147) also peak between 12:00 and 24:00 with a proportionately higher share of incidents during this time.
Incident Volumes By Specific Incident For All Years
100%
149 - Chimney Fires
Incident Volumes
80%
147 - Small Fires
60%
146 - SSC
40% 20% 0% 00:00 to 01:00 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
144 - FAGI
143 - FAM
142 - FDR1
Period
Figure 6.2.2 On the face of this initial analysis moving from Wholetime to Day-Crewing/Retained or other response options will expose a volume of incidents to Retained Response times. Therefore further investigation and analysis needs to be undertaken to define the appropriate service delivery mechanism through the incident response simulator.
Version : 1.3
Page 26 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6.3
Northwich / Winsford / Middlewich Incident Analysis
The following ‘heat map’ shows the total number of incidents occurring per square kilometre over the study period in the Northwich / Winsford / Middlewich area. The BOLD highlighted squares contain the respective Stations Northwich (35), Middlewich (89) and Winsford (42).
10 9 6 5 1 92 69 17 28 42 6 1 3 1 2 3 51 3 3 5 2 1 3 3 8 24 16 21 1 5 213 453 2 6 1 4 1 3 59 283 173 193 12 16 56 3 9 2 80 63 3 33 11 6 17 36 72 114 10 2 6 94 237 76 56 15 56 35 49 197 42 1 2 2 1 3 2 78 159 90 35 31 10 7 197 385 32 6 3 116 9 6 3 4 3 5 4 6 10 41 36 135 151 2 6 15 10 28 12 162 1 3 1 3 6 3 24 15 1 2 1 2 147 89 38 2 2 1 18 5 49 27 26 3 2 4 5 4 5 34 55 7 5 5 3 3 6 6 14 5 1 2 5 1 3 5 5 11 1 3 2 5 3 3 2 7 15 20 26 9 10 8 2 1 8 13 1 1 11 44 1 2 5 5 16 25 18 5 2 1 1 43 1 1
Figure 6.3.1 The following charts show total Incident Volume by Station by time of day and Incident Type by time of day respectively. Small Fires (147) account for over 50% of incidents between 16:00 & 22:00. The use of TRV’s may have a significant impact on the mix of appliances required.
Incident Volumes By Station Ground For All Years
250
Incident Volumes
200 150 100 50 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
Middlew ich
Northw ich
Winsford
Period
Figure 6.3.2
Version : 1.3
Page 27 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Incident Volumes By Specific Incident For All Years
450 400 350 300 250 200 150 100 50 0 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
149 Chimney Fires 147 - Small Fires
Incident Volumes
146 - SSC
144 - FAGI
143 - FAM
142 - FDR1
Period
Figure 6.3.3 It will only be possible to analytically evaluate the impact against Response Standards using the Incident Response Simulator.
Version : 1.3
Page 28 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6.4
Targeted Response Vehicles Analysis
The TRV Response strategy will ultimately be driven by the profile of Small Fire (147) incident demand. It is the response to this demand that needs to be managed effectively to minimise the impact on existing response strategies for other types of incident. Looking at Small Fires (147) by Performance Delivery Group provides a compelling picture of the challenge facing CFRS.
Small Fires (147) by PDG
C hester, 8% W arrington, 21% C ongleton, 3% C rewe and N antwich, 8%
Vale R oyal, 13%
E llesm ere P ort and N eston, 16%
M acclesfield, 6%
H alton, 25%
Figure 6.4.1 Combined incident volumes for Vale Royal, Ellesmere Port & Neston, Warrington and Halton PDGs account for 75% of all Small Fires. This suggests that the greatest return from TRV response options would be obtained by focussing on these four PDG’s. The daily profile of Small Fires (147) incident volumes show demand progressively increasing from 12:00 to a peak at 21:00 before progressively decreasing to 01:00 during a typical day. This represents the outline daily demand profile that a TRV based response option needs to address.
Incident Volumes By Generic Incident For All Years
3000 2500
Incident Volum
2000 1500 1000 500 0 00:00 to 01:00 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
147 - Sm all Fires
Period
Figure 6.4.2
Version : 1.3
Page 29 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
6.5
Aerial / Hydraulic Lift Platform Deployment
The following ‘Heat Map’ shows the geographic dispersion of all incidents for the analysis period across CFRS. In particular it highlights the Unitary Authority’s and heavy industrialised areas in the North / North West of the area and the more rural nature of the South / South East.
Figure 6.5.1 (Included for geographic reference)
Figure 6.5.2 (Heat Map of all incidents - Key as Figure 6.0)
Version : 1.3
Page 30 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Comment [PL2]: Is this all incidents and if so should it just be for incidents requiring a HLP appliance?
If the distribution of HP demand is proportionately distributed across CFRS in a similar manner to all incidents then this suggests an initial 3 base solution as:1) 2) 3) One based in the North West conurbations (Chester or Ellesmere Port) One based in the Northern conurbations (Warrington or Halton) One based in the South East (Crewe or Congleton)
Further in depth study of their actual use over the analysis period needs to be undertaken and analysed. Assuming that HP attendance is more closely aligned to FDR1 incidents and knowing that Chester City incidents require a mandatory HP response the following typical daily profile of Primary Fire FDR1 (142) incidents reinforces the above.
Incident Volumes By PDG For All Years
W arrington
Vale R oyal
800 700
M acclesfield
Incident Volum
600
H alton
500 400 300 200 100 0 00:00 to 01:00 01:00 to 02:00 02:00 to 03:00 03:00 to 04:00 04:00 to 05:00 05:00 to 06:00 06:00 to 07:00 07:00 to 08:00 08:00 to 09:00 09:00 to 10:00 10:00 to 11:00 11:00 to 12:00 12:00 to 13:00 13:00 to 14:00 14:00 to 15:00 15:00 to 16:00 16:00 to 17:00 17:00 to 18:00 18:00 to 19:00 19:00 to 20:00 20:00 to 21:00 21:00 to 22:00 22:00 to 23:00 23:00 to 24:00
C hester C rewe and N antwich E llesm ere P ort and N eston
C ongleton
Period
Figure 6.5.3 The overall hourly demand profile also suggests that a 2 base solution between the hours of 04:00 to 12:00 may be possible. Once again the viability of any combination of locations and number of aerials will be evaluated by the Incident Response Simulator against CFRS Emergency Standards.
Version : 1.3
Page 31 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
7
Analysis – Abstraction Data
At present CFRS maintain paper based records for mechanical downtime, however they are in the process of transferring this information into MS Excel based records to assist in analysing this information. At present electronic records exist from the 1st January 2006 to 30th September 2006. Unfortunately appliances going off the run for a few minutes while the mechanic takes a quick look at it may not get notified to Control and hence are not recorded on the sheet.
Data on non-mechanical downtime primarily related to retained appliance availability is more difficult to provide. Currently this information (Stations telephone control to notify the time and date the appliance will be available / not available for the coming day or couple of days) is entered onto the Remsdaq mobilising system and recorded into an 'Events’ file. This records the time and date the appliance is on/off the run. This then automatically updates the required appliances status to available or not available at the required time. This ensures that for mobilisation purposes Control always know the status of that appliance. Unfortunately once the event is actioned by the system it is automatically deleted from the record.
In order to provide statistical data CFRS currently collate retained appliance availability data in an Excel spreadsheet. This records non-availability during the day or night period but does not specify the exact times of the day. It will merely record for example, that between the hours of 09:00-18:00 an appliance was not available for 4 hours – it will not state which 4 hours of that period. The Service will be introducing within the next month or so, a more advanced electronic availability system for the retained stations - this will record for statistical purposes the exact time of the day when the appliance is available / not available.
A brief addendum report will be produced on abstractions if sufficient data is available before the simulation phase is completed. This will look at :Notification of ‘Off the Run’ Failure to Respond Timing of Abstraction Abstractions by day of week
Version : 1.3
Page 32 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
8
Analysis – 8 Riders vs 9 Riders
This section covers a brief analysis undertaken to look at the difference in Incident Times for 2 Pump attendances between 8 and 9 Rider attendances. This analysis is based on 13,478 incidents observed in the period from 1st January 2003 to 30th September 2006. Incident times have been calculated from the “tstn” (Incident start time) and the “tmrav” (Time appliance mobile returning from incident) fields from the CFRS deployment data. The incident time has been defined as the tstn for the first appliance to be notified to the tmrav for the last appliance to leave the incident. 8 or 9 riders have been calculated by combining the Riders for each appliance. Only 8 or 9 rider incidents have been considered. The following charts show the distributions of 8, 9 and combined 8 & 9 Rider incident distributions. The first displays the raw distributions, whilst the second shows a scale-less version based on dividing by the total number of observations in each population.
Raw Distributions
500 450 400 0.0300 350 300 250 200 150 0.0100 100 50 0 1 17 33 49 65 8 Rider 81 97 113 129 145 161 177 193 9 Rider 8 & 9 Rider 0.0050 0.0000 1 17 33 49 65 81 97 113 129 145 161 177 193 8 Rider 9 Rider 8 or 9 Rider 0.0250 0.0200 0.0150 0.0400 0.0350
Scale-less Distributions
Figure 8.1 The following descriptive statistics provide some insight into these incident time distributions and the average incident times for each population:
This information was used to undertake a statistical t-test into the difference between the means of the 8 & 9 rider distributions using a two tailed independent test.
P value and statistical significance: The two-tailed P value equals 0.9536. By conventional criteria, this difference is considered to be not statistically significant. Figure 8.2 The conclusion of this analysis is that based on average incident times alone there is no difference between 8 Rider attended incidents and 9 rider attended incidents.
Comment [PL3]: Can we reduce the number of decimal places in the table please!
Version : 1.3
Page 33 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
9
Conclusions and Next Steps
This section contains the conclusions that we have drawn from the initial analysis and details the next steps.
9.1
Conclusions
1 : High volume of false alarms Around 40% of overall incidents requiring response turn out to be false alarms. It may therefore be possible make a significant impact on overall demand through initiatives designed to reduce the occurrence of false alarms. 2 : No discernible trend in incident numbers There is no discernible trend in overall incident numbers, although last year saw a rise in Special Service Calls. 3 : Seasonality in incident volumes There is seasonality in incident volumes. The simulator will be able to show whether there is either a need for greater resources in busier months or spare capacity that can safely be reduced in quieter months through expansion of the current strategic reserve. Alternatively, the model may show that it is not possible to reduce capacity in the quieter months due to the need to maintain coverage at all times. 4 : Significant variance in overall incident workload (Small Fires Peak) There is a significant variance in overall incident workload through the day. The Total volume of Incidents per hour is flat between 09:00 and 23:00, however Small Fires (147) can more than double this volume between 17:00 and 22:00 hours. The active use of TRV’s during these periods should be evaluated and encouraged. 5 : Emergency Response Performance CFRS Emergency Response performance against the 2006/07 Emergency Response standards is impressive. At a Brigade level an average of 98% attendance within standard has been achieved over the study period. 6 : 8 vs 9 Rider attendances The conclusion of this analysis is that based on average incident times alone there is no discernable difference between 8 Rider attended incidents and 9 rider attended incidents.
9.2
o o o
Next Steps.
Configure the Incident Response Simulator to use CFRS Incident Data. Calibrate the Simulator against known CFRS performance. Design an experimentation plan to evaluate the five scenarios identified above.
The next steps are as follows:
Version : 1.3
Page 34 of 35
Print Date : 17/11/2006
CFRS : Incident Response Modelling Project
Appendix 1 – Glossary of Terms
This section covers basic definitions and terminology relevant to this report.
Fire Incident Types
Incident Code 142 143 144 146 147 149 151 161 162 190 191 192 194 195 196 197 198 199 Type Cheshire Incident Cheshire Incident Cheshire Incident Cheshire Incident Cheshire Incident Cheshire Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Over the Border Incident Description FDR1 (Structural Property Fires) FAM (False Alarm Malicious) FAGI (False Alarm Good Intent) SSC (Special Service Calls) Small Fires Chimney Fires Shropshire Incident Merseyside Incident Staffordshire Incident Staffordshire Standby Merseyside Standby GMC Standby GMC Incident Northwales Incident Northwales Standby Derbyshire Incident Shropshire Standby Derbyshire Standby
Version : 1.3
Page 35 of 35
Print Date : 17/11/2006
doc_319101559.pdf