Professional Documents
Culture Documents
Recommendation
Staff recommends that Council receive and comment on the attached report from the
International City/ County Management Association (ICMA) and TriData, a division of System
Planning Corporation (SPC).
Executive Summary
TriData, a division of System Planning Corporation (SPC), was contracted, along with the
International City/ County Management Association (ICMA), to conduct a services, resources,
and utilization/deployment analysis of the Palo Alto, CA Fire Department. The attached report
includes not only the data analysis from the city’s computer aided dispatch(CAD) system, but
the findings and recommendations based on that data, as well as information provided by all of
the stakeholders of the study (FD staff, other city staff, City Council, City Administration, etc.).
Background
The scope of this project was to provide a third party independent review of the PAFD such that
the city could better understand how the current system of fire department operations was
working and to determine if the fire department could provide services more efficiently,
effectively, and/or reducing costs in some ways. Specifically, the scope of work required a
focused, objective analysis of overall fire department emergency response operations that
would analyze current resource deployment practices to assess the appropriateness of current
fire department minimum staffing levels in the context of call volume, response times, use of
overtime, deployment practices and employee safety.
Fire department response times (using data from the city’s CAD system);
A copy of the report was provided to the fire department union representatives to share with
fire department personnel the Wednesday prior to tonight’s meeting to gain additional
feedback including the offer to set up a meeting separate from the Council meeting for fire
department staff to discuss the report and its findings with the consultants.
Tonight, the consultants will make a presentation that will include an overview of their
methodology, key findings, and recommendations to the city to improve fire department
operations and deployment of personnel.
Discussion
The comprehensive report recommends a number of changes to fire department structure and
deployment based on actual data from the city’s CAD system as well as modern fire best
practices. In addition, the report highlights the need for some investment by the city in fire
department building infrastructure to achieve some of the recommendations such as a station
merger (see below). Findings are discussed in more detail in the report and some are subject to
meet and confer under existing union agreements, but significant recommendations being
made by the consultants include:
ORGANIZATIONAL STRUCTURE:
Implement a new organizational structure with the current police chief as public safety
director over police and fire and merge the support functions of budget/finance, HR,
IT and planning for police and fire under the public safety director;
Within the merged planning function for police and fire, hire a data/ GIS analyst for the
fire department;
Consider retaining a consultant to assist the city in setting up its ongoing data collection
and analysis process;
Move the senior staff of the PAFD to a single location, preferably with the police
department administration if police and fire are under one public safety director;
Consider adding the rank of lieutenant to replace some existing captain positions –
under this scenario, a captain would be in change of the unit on one shift with
lieutenants supporting (would result in net reduction of captains – through
attrition); and
Combine the emergency management functions of fire and police under the public
safety director – hire a manager to lead this function with qualifications specific to
the field of emergency management – can be a civilian position.
Establish performance goals for dispatch, turnout and travel times and analyze them
monthly and improve turnout times;
Merge Stations 2 and 5 near the intersection of Arastradero Road and Hillview Avenue
(infrastructure improvement needed to accommodate this recommendation);
Eliminate the fulltime staffing of Rescue 2 and cross-staff it and discuss with surrounding
jurisdictions the possibility of sharing the cost of operating this rescue as a ‘regional’
resource;
Eliminate one engine and continue to cross-staff one engine and the rescue, if Stations 2
and 5 are merged; and
Discontinue the overtime staffing of Station 8 on the present schedule and staff the
station only during high fire days as determined by Cal Fire; also consider contracting
with Cal Fire (or another agency), or developing an auxiliary from the existing CERT
members as an alternative; also investigate possibility of using technology to
monitor hot spots in this area and deploy/staff as appropriate.
OTHER RECOMMENDATIONS:
When new fire chief is hired, consider contracting the services of an organizational
development consultant to improve organizational performance, internal
communications, and professional development.
There are also a significant number of recommendations related to improved training and
development of staff, response and turnout times, workload analysis, and staffing changes, as
well as sworn staff to civilian personnel-related recommendations. All of which are described in
detail in the report. These will be covered by the consultants during their report to council at
the study session.
ATTACHMENTS:
Attachment A: Palo Alto CA FINAL REPORT 1-20-2011 pss EDITED II (PDF)
Submitted by:
Philip Schaenman, President
TriData Division, System Planning Corporation
3601 Wilson Boulevard, 5th Floor
Arlington, VA 22201
(703) 351-8300
January 2011
Fire Services Utilization and Resources Study
Palo Alto Fire Department, CA FINAL REPORT
Table of Contents
ACKNOWLEDGEMENTS ..........................................................................................................1
FOREWORD..................................................................................................................................3
EXECUTIVE SUMMARY ...........................................................................................................5
I. INTRODUCTION ......................................................................................................................8
Scope of Project ...........................................................................................................................8
Study Process ...............................................................................................................................9
Organization of the Report ..........................................................................................................9
II. ORGANIZATION AND MANAGMENT ............................................................................11
Major Issues Confronting the PAFD .........................................................................................15
Public Safety Organization ........................................................................................................17
Other Changes............................................................................................................................21
Initial Steps Going Forward .......................................................................................................22
III. POPULATION GROWTH, RISK, AND DEMAND ANALYSIS .....................................24
Data Specification and Issues ....................................................................................................24
Population Growth and Development........................................................................................28
Demand for Emergency Services ..............................................................................................33
Fire Risk Analysis ......................................................................................................................39
Emergency Medical Service Risk Analysis ...............................................................................47
Target Hazards ...........................................................................................................................53
IV. STATION LOCATION AND RESPONSE TIME ANALYSES .......................................55
Strategic Goals ...........................................................................................................................55
Response Time Analysis ............................................................................................................56
Measuring Response Time.........................................................................................................57
Palo Alto Response Times .........................................................................................................60
Workload and Performance Measurement ................................................................................67
Assessment of Fire Station Locations ........................................................................................76
Possible Fire Station Mergers ....................................................................................................82
V. ANALYSIS OF FIRE AND EMS OPERATIONS ..............................................................86
Overview ....................................................................................................................................86
Key Findings and Recommendations ........................................................................................88
Organization and Resources ......................................................................................................90
Emergency Response .................................................................................................................92
Operational Changes ..................................................................................................................93
TriData Division, ICMA Center for Public Safety Management January 2011
System Planning Corporation
Fire Services Utilization and Resources Study
Palo Alto Fire Department, CA FINAL REPORT
TriData Division, ICMA Center for Public Safety Management January 2011
System Planning Corporation
Fire Services Utilization and Resources Study
Palo Alto Fire Department, CA FINAL REPORT
Table 15. Cumulative Distribution Function (CDF) of Response Time of First Arriving Unit
for MVA and EMS Other Calls .............................................................................................. 144
Table 16. Average Response Time for Structure Fire and Outside Fire Calls by First Arriving
Fire Unit ................................................................................................................................. 145
Table 17. Average Response Time for Structure Fire and Outside Fire Calls by Second
Arriving Fire Unit ................................................................................................................... 146
Table 18. Cumulative Distribution Function (CDF) of Response Time of First and Second
Arriving Fire Units for Structure and Outside Fire Calls ....................................................... 148
Table 19. Frequency Distribution of the Number of Calls ..................................................... 150
Table 20. Top 10 hours with the most calls received ............................................................. 151
Table 21. Unit Workload Analysis Between 8 a.m. and 9 a.m. on 17-Feb-2010................... 152
Table 22. Unit Workload Analysis Between 9 p.m. and 10 p.m. on 19-Sep-2009 ................ 154
Figures
Figure 1. Non-canceled Calls by Type and Duration ............................................................. 120
Figure 2. EMS and Fire Calls by Type ................................................................................... 122
Figure 3. Calls by Month ........................................................................................................ 123
Figure 4. Calls by Hour of Day .............................................................................................. 124
Figure 5. Number of Units Dispatched to Calls ..................................................................... 125
Figure 6. Busy Minutes by Hour of Day ................................................................................ 131
Figure 7. Average Dispatch, Turnout, Travel, and Response Time of First Arriving Units by
Call Type ................................................................................................................................ 137
Figure 8. Number of Total Calls for the First Arriving Unit .................................................. 139
Figure 9. Average Dispatch, Turnout, Travel, and Response Time of First Arriving Unit by
Hour of Day ............................................................................................................................ 141
Figure 10. Cumulative Distribution Function (CDF) of Response Time of First Arriving Unit
for MVA and EMS Other Calls .............................................................................................. 143
Figure 11. Cumulative Distribution Function (CDF) of Response Time of First, Second and
Third Arriving Fire Units for Structure and Outside Fire Calls ............................................. 147
Figure 12. Unit Workload Analysis by Call Type Between 8 a.m. and 9 a.m. on 17-Feb-2010
................................................................................................................................................ 153
Figure 13. Unit Workload Analysis by Call Type Between 9 p.m. and 10 p.m. on 19-Sep-2009155
TriData Division, ICMA Center for Public Safety Management January 2011
System Planning Corporation
Fire Services Utilization and Resources Study
Palo Alto Fire Department, CA FINAL REPORT
TriData Division, ICMA Center for Public Safety Management January 2011
System Planning Corporation
Fire Services Utilization and Resources Study
Palo Alto Fire Department, CA FINAL REPORT
TriData Division, ICMA Center for Public Safety Management January 2011
System Planning Corporation
Fire Services Utilization and Resources Study
Palo Alto Fire Department, CA FINAL REPORT
ACKNOWLEDGEMENTS
We wish to thank the members of the Palo Alto Fire Department (PAFD), city and police
department officials, and other contributors that provided valuable insight for this study. They
not only provided information and welcomed our visits to their areas but also interacted helpfully
with us throughout the study.
Acting Public Safety Director Dennis Burns and Deputy Fire Chief Roger Bloom along
with other career officers, firefighters, EMS personnel and support staff members were
extremely cooperative, gracious, and forthcoming during the project. They provided data,
facilitated tours of fire stations, and they discussed openly the challenges and opportunities
facing the organization.
Although we received valuable input from a variety of sources, the findings and
recommendations contained in this report are entirely those of the TriData and ICMA project
team. Principal members of the team and their areas of responsibility are shown below.
TriData Staff
Philip Schaenman Corporate Oversight
Stephen Brezler Project Manager; Fire & EMS Operations
Markus Weisner Risk and Demand; Station Location Analysis
John Montenero Organization and Management; Service-Sharing Opportunities
Peter Moy Staffing, Overtime and EMS Cost Review
Maria Argabright Production Coordinator and Project Support
FOREWORD
TriData, a division of System Planning Corporation (SPC), is based in Arlington,
Virginia. SPC is a 200-employee defense and national security contractor specializing in high-
level systems engineering and national security.
Over the past 29 years, TriData has completed over 250 fire and EMS studies for
communities of all sizes. In addition to this local government consulting, TriData undertakes
research in a wide range of public safety issues. TriData is also noted for its research on behalf of
the United States Fire Administration (USFA), Department of Homeland Security (DHS), and
other Federal and state agencies, as well as the private sector. TriData also conducts international
research on emergency response topics and has conducted extensive research on effective fire
prevention strategies in Europe and Asia.
_________
The International City/County Management Association (ICMA) is the premier local
government leadership and management organization. Since 1914, ICMA‘s mission has been to
create excellence in local governance by developing and fostering professional local government
management worldwide. ICMA provides an information clearinghouse, technical assistance, and
training and professional development to more than 9,100 chief appointed administrators,
assistant administrators, and other individuals throughout the world. The organization's resources
and services reach thousands of local, state, and federal government personnel, academics,
private sector professionals, citizens, and other individuals with an interest in effective
management at the local government level.
to provide technical assistance to more than 70 communities during the past two years. The
Center also undertakes research, participates with other organizations and governments to share
best practices, conducts training and education, and assists ICMA International in developing
countries. The Center works with ICMA Publishing to document practices and educate local
government leaders. Ongoing assistance is provided to the National League of Cities and
Department of Homeland Security for education and training of elected as well as appointed
officials.
EXECUTIVE SUMMARY
TriData, a division of System Planning Corporation (SPC), was contracted, along with
the International City/ County Management Association (ICMA), to conduct a resource
deployment analysis of the Palo Alto, CA Fire Department. The Palo Alto Fire Department
(PAFD) protects over 77,000 residents, a major university (Stanford), a research facility
(Stanford Linear Accelerator), and the Palo Alto foothills. To accomplish its mission, 105
personnel are assigned to staff eight stations. Seven engines, one ladder, one rescue and one
advanced life support unit (ALS) are deployed to the eight stations 24/7. A second ALS unit is
operated on a part-time basis as is a basic life support (BLS) transport unit staffed weekdays by
civilian EMTs. During the high fire season, July to November, a wildland suppression unit is
also staffed.
Implement a new organizational structure with the current police chief as public
safety director over police and fire;
Hire a data/ GIS analyst;
Move the senior staff of the PAFD to a single location, preferably with the police
department administration;
Discontinue the practice of developing entrance and promotional exams within the
fire department;
Combine the emergency management functions of fire and police under the public
safety director.
Data collection is a significant issue for the Fire Department, which is using an older
version of the National Fire Incident Reporting System (NIFRS) and there are questions of data
quality and completeness.
From 2000 to 2009, total calls increased only 19 percent but medical calls increased by
48 percent. The trend analyses determined that total calls are likely to continue to increase
slightly through 2020, with medical calls adding most of the demand. Fires, in particular
structure fires, are infrequent occurrences. Data showed that Palo Alto has a lower fire death rate
per capita than the US, the west coast, and communities of similar size. The average dollar loss
per fire incident is also lower than the US and communities of similar size, despite its relative
affluence.
Total response times, which include the three time segments of call processing (dispatch),
turnout (reaction time), and travel time, are generally good. However, turnout times are too long
and dispatch times were so low that it appears that the call-taking portion of the time segment is
not included on the time stamp. Travel times for the stations ranged from 4:30 for medical calls
to 5:22 for hazardous condition calls. Travel times are slightly higher than the recommended
metric of 90 percent arrival within four minutes. Station 4 (Middlefield) has the highest average
travel time (5:18) and Station 2 (Hanover) the lowest (4:04).
Establish performance goals for dispatch, turnout and travel times and analyze them
monthly; improve turnout times
Merge Stations 2 and 5 near the intersection of Arastradero Road and Hillview
Avenue;
Discontinue the practice of mandating the minimum number of on duty personnel
within the MOU;
Eliminate the fulltime staffing of Rescue 2 and cross-staff it. Discuss with
surrounding jurisdictions the possibility of sharing the cost of operating this rescue as
a ‗regional‘ resource;
Staff a second medic unit 24/7;
Eliminate one engine and continue to cross-staff one engine and the rescue, if Stations
2 and 5 are merged;
Discontinue the overtime staffing of Station 8 on the present schedule; staff the
station only during high fire days as determined by Cal Fire; also consider contracting
with Cal Fire (or another agency), or developing an auxiliary from the existing CERT
members as an alternative; and
Discuss with Menlo Park the possibility of sharing the cost of operating Station 1 in
Palo Alto with Station 7 in Menlo Park.
This study also reviewed the fire prevention activities and training. Recommendations to
improve these areas include:
Fill the vacant fire marshal position and transition the fire code enforcement division
to a civilian (non-safety) staff; also add a public education specialist
Conduct a complete audit of all training requirements needs and prepare a multi-year
plan that includes staffing requirements and cost;
Implement a formal program of career and professional development, including
requirements for promotion for all officers;
Conduct more training during evenings and weekends; and
Regionalize the city‘s fire and EMS training program and plan for the replacement of
the city‘s training facility.
I. INTRODUCTION
TriData, a division of System Planning Corporation and its partner consultant on this
project, the International City/ County Management Association (ICMA), were selected to
conduct an independent review of the City‘s emergency services, primarily fire and EMS.
The Palo Alto Fire Department (PAFD) protects a combined population of just over
77,000 residents in a uniquely complex area that includes a major university (Stanford), a
research facility (Stanford Linear Accelerator), and the Palo Alto foothills (25 square miles of
mostly open woodlands). To provide emergency services, the PAFD is deployed from eight fire
stations. Operating from these stations 24/7 are seven engines, one ladder truck, one rescue, one
ALS rescue, and one battalion chief. The fire department also provides one ALS rescue (12
hours per day) and a smaller engine (Type III) for the foothills area by assigning fire personnel
on overtime. The current service-delivery model requires a minimum on-duty staff of 30
personnel.
Usually managed by a fulltime fire chief, at the time of this study the police chief was
acting as the city‘s public safety director and was responsible for the executive-level
management of the fire and police departments. A deputy chief manages the fire department‘s
day-to-day activities.
Scope of Project
The scope of this project was to provide a third party independent review of the PAFD
such that city officials can understand how well the system is working. Officials also desired to
understand whether the fire department can provide services more efficiently. Specifically, the
city requested ― a focused, objective analysis of overall fire department emergency response
operations that will analyze current resource deployment practices to assess the appropriateness
of current fire department minimum staffing levels in the context of call volume, response times,
and employee safety‖.1
Fire department response times (using data from the city‘s CAD system)
Fire and EMS unit workloads
Services and agreement with Stanford University
Fire station and staffing locations/ concentrations
1
Request for Proposal, Attachment C, Part I.
Study Process
The process used for this project combined multiple research techniques including personal
interviews, collecting and reviewing background information, analyzing CAD and incident data, and
performing geographic information system (GIS) analyses. We also visited each of the city‘s fire
stations and we drove throughout the city to understand its unique setting and geography.
TriData‗s project team conducted a series of meetings with local fire and government
officials. From the beginning and throughout the project, TriData coordinated its efforts through the
ICMA project manager since the ICMA was conducting a parallel study of the city‘s fire and EMS
demand. We also discussed with the ICMA our ideas and rationale for the major findings and
subsequent recommendations in this study. Throughout the project we maintained contact with the
city‗s designated project manager and we followed up with key individuals by e-mail and telephone.
Chapter III, Population Growth, Risk and Demand Analysis – This chapter discusses
the pressures on the fire system, including an analysis of future population changes and a
projection of demand. These factors are important in evaluating future viability of the system and
identifying resource needs related to future demand for fire and EMS services.
Chapter IV, Station Location and response Time Analyses – This chapter presents the
geographic information system (GIS) analysis of station locations and response times. The
section discusses the station configuration recommended by this study and it includes the maps
depicting the current and proposed system.
Chapter V, Analysis of Fire and EMS Operations, Station Locations and Response
Times – This section discusses fire suppression, including discussions of the command structure,
staffing, and the deployment of apparatus and staff. Presented here are the major deployment
modifications recommended. A discussion of overtime and the staffing analysis is also in the
chapter.
Chapter VI, Fire Prevention, Training and 9-1-1 Dispatch – This chapter discusses
the delivery fire prevention and code enforcement which are so important to risk management.
The PAFD training division and the operation of the city‘s 9-1-1 dispatch system is also
presented.
The PAFD is organized similar to other fire departments; albeit it is unusual (but not
unique) in providing protection to a major university in a service-sharing arrangement. In 1976
the city and Stanford University merged their fire departments and two of the Department‘s eight
stations are located on University property. Under the merger, Stanford University funds 31
percent of the department‘s total budget, including operating and capital outlays.
The fire department has three primary divisions with a total staff of 121 full time
equivalent (FTE) personnel. There are also part -time (hourly) positions that equate to another
4.4FTE positions. The part- time positions are civilian (non-safety) positions used to staff a Basic
Life Support (BLS) ambulance on weekdays to handle inter-facility transports. Divisions within
the PAFD include emergency response, environmental and safety, training and personnel.
Authorized FTE personnel for the PAFD, not including the part-time BLS personnel are
shown in Table 1.
Table 1: PAFD Authorized Positions, 2010
Fire personnel are represented by the International Association of Fire Fighters Local
1319. The current four-year MOU expired on June 30, 2010 and at the time of this study the city
and union were negotiating a new agreement. Chief Officers are represented by the Palo Alto
Fire Chiefs‘ Association. Chief Officers are also in negotiations as their contract also expired on
June 30, 2010.
The FY2010 fire department budget is $26.3M. Because Stanford University pays 31
percent, the taxpayers are getting excellent value due to their lower cost per capita for emergency
services. The city could not reduce its fire budget by 31 percent if Stanford were not part of the
system because most of the present cost of operating the fire department would still be necessary
since it is unlikely the city could eliminate both Stanford stations if it did not protect the
University. Also the city would still need to cover the full cost of the administrative staff, its fire
chief, at least one deputy chief, and three battalion chiefs. The full salary (and benefits) for these
positions, and probably others, would be paid 100 percent by the city if it were not partnering
with Stanford.
Going forward it is in the best interest of the city to maintain its efficient service-sharing
relationship with Stanford and it should consider the University a valuable partner. As the city
considers future options on how to provide emergency service, it should also consider that cost
savings are important to the University as well, and consider ways to lower the University‘s cost.
Table 2 depicts the 2010 fire department budget. Of the 2010 budget, salaries and
benefits accounted for $21.6M or 82 percent of the fire department‘s total budget. Percentage
wise this is slightly lower than other fire departments where salaries and benefits typically
average from 85-92 percent of the total budget.
Table 3 shows the salary for the various positions within PAFD. These were used as the
base point from which to estimate the savings or additional costs if the city implemented any of
the deployment changes recommended later in this study.
Table 3: Fire Department Salaries, 2010
The MOU now in effect requires employees to be paid for working out of classification
or maintaining certifications such as paramedic. Fire captains working a 40-hour week are also
paid a premium rate even though the hours worked per week are less than for captains assigned
to shift work. The rationale is usually that shift work is preferred over a weekday schedule
because while on shift work fire captains earn FLSA pay. Captains also lose holiday pay when
on the 40-hour schedule. Pay for working out of classification and maintaining certifications
above those for other employees in the same classification are common practice in the other fire
departments.
Table 4 shows the pay differential for the professional qualifications within the current
MOU.
Table 4: Pay Differential Schedule
The fire chief‘s position is currently vacant. A deputy chief‘s position (training and
personnel management) is also vacant and the position is being filled part time by a retired
deputy chief. And there are other officer positions such as station captain that are currently
unfilled. Of the fire department‘s key management team, five senior positions are either unfilled
or have individuals acting out of classification. This situation is contributing to the fire
department‘s inability to be a forward looking organization. Positions unfilled or with
individuals in an acting capacity include:
Over the years however, the fire department has become a stagnant organization. While it can
still be proud of its ability to keep the public safe when an emergency occurs, management of the
department is struggling, in part because there are few support staff to conduct important
planning activities. Additionally there is an insistence within the organization, even within
management, to focus most of its attention on fire suppression even when emergency medical
calls are the majority of the Department‘s business, and prevention is a key to reducing losses
and reducing demand (calls).
The labor approach in Palo Alto (and nationally) has been a strong aggressive push to use
all available means to protect the current system and to enhance services such that firefighter
numbers are increased by placing restrictions in their labor agreements. In California, passage of
the Meyers-Milias Brown Act establishing meet and confer requirements for public employees
were the beginning of these efforts. The current MOU has staffing requirements that specify the
minimum number of personnel to be on duty. As a result, the city‘s ability to determine the
desired level of service is very much restricted. Changes to the deployment system in Palo Alto
have been difficult, if not impossible to achieve.
Even so, those with whom we spoke reported that labor-management relations between
the city and firefighters were generally good. For the city going forward, any hopes at changing
resource deployment for fire and EMS services will require modifications to language in the
MOU. For example, the fire department is currently required to staff its ALS rescues with two
paramedics even though one paramedic and one EMT is acceptable under the protocols of the
County‘s Health Officer, which regulates EMS service standards. To change EMS staffing to one
paramedic and one EMT, a change wanted by the fire department management staff, requires a
change to the MOU, or at least a side-letter agreement.
A leadership malaise has occurred within the fire department over the past few years,
part of which is due to poor training for management staff
There is a an absence of relevant data to determine if the department is providing the
right services, at the right places, at the proper level
Firefighters though dedicated, spend little time on prevention activities and training
after their initial recruit training is poor
Officers get little training and there are low expectations for their performance
Separate OES operations within the police and fire departments are not coordinated
effectively2
Battalion chiefs, who are important in policy implementation, each have assigned
collateral responsibilities such as coordinating fleet maintenance and other support
activities. However, not all of them take their extra responsibilities seriously.
Reductions in the fire department‘s budget over the past few years have eliminated
important staff positions
2
As an example, all employees are to be trained in the National Incident Management System (NIMS) but
reportedly not all are receiving the training. Likewise, some city departments are reportedly not cooperating with
OES at the level necessary to achieve an effective outcome in a major emergency.
Fire management recognizes many of these issues and several commented about the
issues the department must prepare to address. Among them were:
A large turnover in fire department personnel is likely over the next few years as the
expected number of retirements increases
Apparatus, especially ambulances that are supposed to be replaced at four years, are
now over 10 years
Stations 3 and 4 are in need of major upgrades
The issue of adequate support staff has become critical
$1.1 Million was budgeted for overtime each year for the past 10, but overtime
continues to increase beyond that level12-hour medic unit was operated using
overtime for the past 21 years
Loss of key staff support means that the PAFD is only able to plan and address major
issues by using voluntary committees
Reductions and deferrals to capital projects must be addressed as the fire departments
facilities have become outdated and in some cases too small
Although this study did not include a review of the Palo Alto Police Department (PAPD),
out of necessity we met with police officials because the fire dispatch center is managed by the
police department. We also toured the dispatch center and met with various officials within the
PAFD. The fire chief‘s office and fire prevention offices are located at city hall as is the entire
PAPD operation.
A major problem in the fire department is management support. Functions such as
planning, data analyses, budget and finance, HR and IT all need major improvements. For the
police department, these are areas of strength. Like most police departments the PAPD relies on
data and planning to determine how best to deploy its resources.
Although the city may have intended the public safety director‘s position as a temporary
move, it should consider making it a longer-term arrangement and merge the support functions
of the police and fire departments. The fire department‘s need for improved staff support can be
achieved if the departments were merged under a single director and the change would make the
operation of both agencies more efficient because of the shared support functions. The interim
director is a solid choice because he is well regarded within city government and the fire
department. He also understands the city‘s politics.
Figure 2 and Figure 3 show the current PAPD organization and how a combined
organization under a public safety director might look.
Figure 2: Current PAPD Organization
Police Chief
Administration
Personnel Services
(Assistant Chief)
Animal Services
PAPD
FTE Positions: 154.5
Figure 3: Potential Organization-- Police and Fire under a Public Safety Director
City Manager
Public Safety
Director
Under a revised fire department structure as part of a combined public safety organization
with a dedicated director, there are several options. The city could eliminate the fire chief‘s
position and instead have a chief of operations equivalent to that of a deputy fire chief. Another
option is to retain the position of fire chief and eliminate one of the two deputy chief positions.
Either can be effective and would achieve some savings. That one of the two deputy chief
positions is already part time and the fire chief position is unfilled makes this a good time to
make such a change.
Recommendation 1: Implement a new organizational structure with the current police chief
as public safety director. At the same time, merge the support functions of budget/ finance,
HR, IT, and planning for police and fire under the public safety director.
If, as we recommend, the city decides to merge fire and police support functions, there is
also a need to improve data and GIS analyses within the PAFD. From our work on this project it
was evident that the fire department does not have the technical acumen in-house to analyze its
activities, response times relative to incident outcomes, and other important criteria. For the fire
department to build on the recommendations of this study and analyze possible changes beyond
those of this study, it should be able to conduct detailed analysis. To do so will require expertise
to evaluate fire loss, patient outcomes, and other activities related to the department‘s response
activities.
Recommendation 2: Within the merged planning function for police and fire, hire a data/
GIS analyst specifically for the fire department.
As was mentioned the fire department has little experience with data analysis. For this
project we spent an inordinate amount of time getting the CAD data to match with incident
reports. The reason is that the two data sets often had different incident numbers and addresses
for the dame incident. As part of the process to improve data collection for management use, it
would benefit the city to have a consultant work with city staff on the issues of fire and EMS
data collection, and then set up the arrays such that analysis is straightforward.
Recommendation 4: Move fire management and the senior staff of the PAFD to a single
location, preferably along with the police department administration if police and fire are
merged under a public safety director. Leasing space may be preferred to constructing a new
facility.
Just as location is hindering the fire department‘s effectiveness, so too are the
expectations of managers throughout the entire department. Though not totally clear as to why,
the expectation of officers is not very high in the PAFD. In many departments battalion chiefs
and even captains have significant responsibilities and they have specific goals to accomplish,
typically within those of the department‘s strategic plan. In some ways the low expectation of
officers is emblematic of the fire service overall because fire officers are promoted from within
the ranks and then are required to manage the labor group from which they were part.
A result of the location problem and not having all chief officers (and captains) fully
committed to solving problems and offering new solutions is that all issues eventually find their
way to top of the food chain where the fire chief or a deputy chief becomes overburdened. This
situation is occurring in the PAFD.
Recommendation 6: Consider adding the rank of lieutenant. Under this scenario a captain
would be in charge of the unit on one shift with lieutenants on the other two shifts. This
change would result in a net reduction of captains (through attrition) and approximately a ten
percent savings per FTE position as each captain is replaced by a lieutenant. Span-of-control,
supervisory oversight, and professional development are also improved.
Other Changes
For a long time the fire department has prepared and administered its own promotional
exams and intake processes for new firefighters. This is a highly unusual practice and few fire
departments have the skill set (and time) to perform such a major HR function. A deputy chief,
who by all accounts is very knowledgeable about the promotional test and requirements, is the
lead individual for the PAFD.
Taking on such an activity within an already short-staffed management team does not
seem to be a good use of resources. Also, developing and administering promotional
examinations and recruit testing are two areas that subject the city to increased legal exposure.
For these reasons, many cities either have their central HR division administer exams, or they
contract these services to a qualified vendor. The city might also consider improving the
efficiency of the testing process by partnering with surrounding communities to standardize the
process since they too have the same needs.
Recommendation 8: Combine the emergency management functions of fire and police under
the public safety director. Hire a person as the manager with qualifications specific to the field
of emergency management. The current coordinators in police and fire can be support staff to
the EM director; However, it is suggested that the current fire position coordinating EM be
changed to a civilian.
To avoid such a situation in Palo Alto, the city should consider using an outside
consultant to begin the implementation process of this study‘s major initiatives while at the same
time assisting the city in the selection of its new fire chief (or chief of fire/ EMS operations,
whichever the case might be). In so doing the city can begin to implement the required changes
within the fire department. The added advantage is that potential candidates understand the city‘s
long-range plan for the fire department and the candidates can be assessed as to their
understanding of the city‘s plan. New ideas that support the city‘s plan can also be solicited from
the candidates.
Recommendation 9: At the initial stage of a police and fire functional support merger,
contract for implementation services to oversee the process and assist with the selection of a
new fire chief, who then would report to the director of public safety.
Recommendation 10: When/if a new fire chief is hired, consider contracting the services of
an organizational development consultant to improve organizational performance, internal
communications, and professional development.
Although Palo Alto does have designated neighborhoods, they are not used extensively
for planning purposes. For this study it was difficult in some cases to place a particular incident
within the correct neighborhood. However the results do appear to be close enough to understand
the emergency call situation for the city as it relates to its various communities. Going forward
the fire department can improve its analysis and use these neighborhoods for planning purposes.
NFIRS Data Issues – The National Fire Incident Reporting System (NFIRS) is a system
established by the National Fire Data Center of the United States Fire Administration (USFA) to
gather and analyze information on the magnitude of the Nation's fire problem, as well as its
detailed characteristics and trends. The first version of NFIRS released in 1975 only collected
fire incident data on a paper-based form. Over the last 35 years, the system has progressed to
version 5.0 and now includes EMS, hazmat, and other data collection modules to reflect the all-
hazard nature of current-day fire department work.
Although Palo Alto Fire Department does collect NFIRS data, it is still using an older
version of the system and there were questions of data quality and completeness. The fire
department provided us data from 2004 through the present, but the data prior to 2007 was
missing approximately 80 percent of incidents. The fire department staff worked with Zoll Data
Systems, their NFIRS software vendor, to troubleshoot the problem, but no solution was found.
Ultimately, staff at the dispatch center had to tap into the underlying database structure to
provide us raw data for fire department incidents from 2000 to present.
Recommendation 11: Upgrade NFIRS data collection to version 5.0 and establish a process
by which the fire department can readily access NFIRS data for conducting analysis.
Recommendation 12: Continue to work with Zoll Data Systems to fix 2004-2007 incident
records. Historical data plays an important role in fire department strategic planning and it is
important that PAFD has accurate and complete NFIRS data.
CAD Data Issues – Computer Aided Dispatch (CAD) data provides a permanent record
of communications between the dispatch center, the public, and fire department units in the field.
Typically CAD data includes an incident dispatch type, incident address, incident coordinates,
and response timestamps for all responding units. Although we received great cooperation and
help from the dispatch center technical staff, we identified some critical data issues.
The current Palo Alto CAD system is outdated and does not track incident coordinates.
Newer CAD systems allow call-takers and dispatchers to display and verify incident locations in
real-time. These systems record an accurate GPS coordinate for each incident (assuming
dispatchers correctly verify incident locations). Palo Alto is still using an older system where
units are picked based on an address database rather than real-time address mapping. In this
system, call-takers have to either input addresses that perfectly match up with the database or
manually figure out where an incident is located and what units to dispatch. This type of system
has no way of determining or recording GPS coordinates and, as a result, there is no definitive
incident location recorded. This type of system limits the accuracy of neighborhood-level fire
department analysis that depends on knowing where all incidents are located.
Recommendation 13: Upgrade the CAD system so that GPS coordinates are collected for all
incidents.
Beyond the lack of coordinates, there are some inherent problems in using CAD data for
conducting performance measurement. CAD data typically contains many inaccurate records. It
is rare to find fire departments that actually analyze their data on a regular basis and, as a result,
there is often no driving force for insuring data accuracy. The CAD data we received from the
Palo Alto dispatch center contained many incidents that were missing unit timestamps, had
clearly incorrect unit timestamps or showed duplicate unit responses on the same incident. To
correct this problem, a better system needs to be put in place for accurately recording fire
department performance.
We recommend putting in place a data quality program encourages the fire department
staff to verify and address data inaccuracies. CAD data is typically transferred over the NFIRS
software/database where line-staff and supervisors can review and correct data. A captain
entering incident data should notice if the incident shows incorrect timestamps or apparatus and
can work with supervisors and/or the dispatch center to correct these inaccuracies before they
become part of the permanent data record. Regular monthly data analysis reports distributed to
the entire department can also identify potential inaccuracies in the data collection process.
We did not have a chance to conduct an in-depth review of the CAD/NFIRS interface,
nor did we get into the details of how the fire department reviews, verifies and corrects incident
data. Although the fire department probably does have some data review procedures in place, the
CAD and NFIRS data issues we identified in this study highlight the need for updating and
improving the entire process.
Recommendation 14: Work with dispatch staff, the CAD software vendor, the NFIRS
software vendor, and line-staff to put in place a process by which the final incident data
entered into the database contains no gross inaccuracies.
Planning Area Issues – It is good practice for fire departments to consider risk and
demand at a neighborhood level. In order to accomplish this, fire departments must have some
predefined planning areas by which to summarize incident information. It was apparent during
our study that Palo Alto did not have any planning areas that could be used for fire department
analysis. In the Land Use section (chapter 2) of the Palo Alto Comprehensive Plan, it states that
―areas are classified as Residential Neighborhoods, Centers, or Employment Districts‖ and that
―understanding the linkages and connections between these areas and within the region is critical
to integrating land use and transportation planning.‖ These areas are, however, not particularly
conducive to conducting fire department planning because many of the ― planning areas‖ are only
a couple blocks large to encompass individual development projects.
Palo Alto also has a set of historical neighborhoods and political neighborhoods, but
neither of these neighborhood sets covered the entire Palo Alto Jurisdiction. Ultimately, we had
to create our own planning areas for this study. We used Palo Alto‘s 24 historical neighborhoods
and added the Stanford University Campus, the commercial area around Stanford shopping
center, and the Bayshore area.
Note: When we refer to planning areas in this report, we are referring to these 27 areas
rather than the Palo Alto Comprehensive Plan ― areas.‖ Figure 4 shows the 27 planning areas
used for our neighborhood-level risk and demand analysis in this chapter.
\\
...."".
~
I
1
.. u.. ...
Palo Alto, CA
• PaIoAho Fir. Station.
Geocoding Issues – As was already mentioned, the current Palo Alto dispatch center
CAD system does not record GPS coordinates for each incident. As a result, we had to attempt to
geocode all incidents based on their address. This process is never perfect because of incorrectly
entered addresses or missing incident addresses. Palo Alto does have a significant problem of not
being able to match incident geocodes to correct addresses. Thus for this study we geocoded
three years of data (2007-2009) to achieve a better match rate (97 percent). This means that only
three percent of incidents were not located and, thus, not included in the neighborhood-level
analysis.
From 1997-2008 there was a large amount of development taking place in Palo Alto,
which manifested itself as a population increase in the late 2000s, just before the recent
economic downturn. The Association of Bay Area Governments (ABAG) predicted a 2020
population of 70,400 (14.4 percent growth) largely based on the very recent period of accelerated
growth, but Palo Alto planners do not believe this prediction to be accurate. Based on their
limited land availability and limited redevelopment potential, Palo Alto developed an in-house
set of population predictions. Both the ABAG and Palo Alto projections are shown in Table 5.
Based on our discussions with Palo Alto planners, we are assuming the Palo Alto projections are
a more accurate prediction on which to base fire department incident trends.
4
Figure 5: Actual (solid black) and Predicted (dashed green) Population
3
City Manager Report to Council; Comprehensive Plan Amendment and Housing Element Update, March 3, 2010.
4
Ibid.
Daytime Population – Residential population provides only a part of the total growth
picture because of the impact that jobs and commuters have on the non-residential / daytime
population. This component of growth is an especially important consideration for Palo Alto
where a large number of commuters access the downtown and university avenue region during
business hours.
The Palo Alto Cal Train station is second busiest station on the line because of all the
commuters coming into the Stanford Research Park, the downtown area, and the California
Avenue area.
Palo Alto, CA
• Pa lo Alto Fire St ations
E::I St anford Boun dary
[=] Pa lo Alto Boundary
2009 Daytime to Nightime Population Ratio
• 66t03,08 0,100
• 51 to 65
36 to 50
21 to 35
o to 20
Zero Pop ulation
It can be seen that Stanford University, the eastern portion of Downtown North,
University Park, the western portion of University South, the research park, large parts of Green
Acres, Charleston Terrace, and most of the neighborhoods just south of Alma Street all have
significant daytime populations. This is significant because demand is affected by large daytime
population increases, a fact that is relevant in planning the deployment of resources, especially
for EMS.
Expected Development – As will be discussed in the EMS risk section of this chapter,
the number of EMS incidents is almost directly proportional to population. Understanding the
expected population changes before they occur help the fire service to prepare for changes in
service demand. Furthermore, it is necessary to plan adequate fire service protection for new
buildings. New development and growth areas always need to be considered in the evaluation of
fire station and apparatus locations.
Palo Alto currently has no vacant land inventory and the planning department stated that
the urban area of the city is fully built out. Although the city does have additional vertical
capacity, it is questionable how much vertical development will actually occur. Buildings have a
max height of 50 feet downtown, 30-35 feet in the El Camino Real area, and there are a couple
other areas that are zoned for higher density growth and building heights up to 50 feet.
A big project on the horizon is the Stanford Medical Center expansion that will account
for an additional 2,000 jobs. This project is not expected to be completed until after 2020.
There is also a big push to build more houses because of all the businesses in Palo Alto,
but equal pushback from the community about growth because of schools and infrastructure
concerns. The planning department predicts an average annual increase of 100 housing units
going forward, which will have minimal impact on EMS. It is expected that the California
Avenue area will continue to grow and that the commercial area along El Camino Real will
continue to see increases in residential population because the zoning allows for mixed-use
commercial/residential projects. Planners also expect some additional development in the
downtown area.
Because of the current economy and the fact that Palo Alto is pretty much built-out, there
is very little on the development front that will have a large impact on the fire department. The
fire department should, however, keep in close contact with the planning department and work
with them to ensure that any large developments will have adequate emergency services
performance.
Fire Unit Locations – Planning areas with high levels of demand show where fire
apparatus should be located. Further understanding of the types of incidents in each
area helps to determine the type of response equipment is most appropriate.
Weight of Response – The demand analysis shows the number of fires versus the
number of fire alarms. A higher weight of response is prudent for planning areas with
higher-risk properties, and where more structure fires actually occur.
Prevention – Some areas have such a high demand for emergency service that
prevention and education efforts must often be increased.
Jurisdiction-Wide Incident Type Trends and Predictions – Using a statistical
software package, a multi-linear regression procedure was used to investigate how both time and
population affect each of the individual incident types. (Time reflects changes in inclination to
use EMS and factors other than population growth per se.) A best-fit multi-linear model was used
to analyze each incident type and the population projections (discussed earlier) and these were
then used to predict future demand. Certain incident types, such as EMS, are much more
dependent on population than others. This is all built into the multi-linear regression model.
Ideally, we would include information about housing stock, population demographics and
other variables in our model. Although many of these additional statistics are available as
historic data, they are usually not projected by jurisdictions. We would need both historical and
projected values to use within our model. For that reason, we have chosen a simpler model that is
only based on two inputs.
For any model, it is necessary to say how statistically accurate it is, or what the
confidence is in the estimate. s. For example, if we predict that there will be 200 fires ten years
from now, we also have to state the confidence limits of that prediction. The confidence interval
is a statistical plus/minus calculation. To continue with the fire example, we might say there will
be 200 fires, plus or minus 50. This gives the reader both a prediction and a range within which
we are fairly certain (95 percent certain to be exact) that the eventual total will fall.
The confidence intervals are the result of a statistical calculation that analyzes how
accurately our prediction model represents the actual data. A very good model will have a small
confidence interval. Our "Total Incidents" predictions usually have small confidence intervals
because the annual totals stay fairly steady and trends typically remain uniform from year-to-
year; as a result a multi-linear regression is able to make fairly accurate predictions for total
incidents for at least several years into the future. The further into the future, the wider the
confidence limits.
Large confidence intervals occur when there are big incident type fluctuations from year-
to-year that are inconsistent and cannot be accurately modeled with any of the independent
variables (time and population). For instance, if the annual number of fires fluctuates up and
down 30 percent from year-to-year, the model cannot accurately predict the exact number of
fires for a given year. In that case, there would be a large confidence interval that essentially says
we predict X, but the number could be much higher or much lower.
When combining large confidence intervals with incident types that already have low
annual totals (e.g. fires), it is not uncommon to see confidence intervals that drop to zero.
Although we know it is unlikely fires will drop to zero, our model may not be able to statistically
confirm that. It is for that reason that we have to combine judgment with these statistical models
to ultimately decide what direction we expect a particular incident type to trend. Our statistical
projections are by no means perfect, but rather tools to help understand incident type trends and
identify potentially extreme increases or decreases of emergency demand.
The figures on the next page show the trends by incident type and projections for 2010
through 2020. The black line that stretches from 2000 to 2009 represents actual incident totals
for those years. The red line that stretches from 2009 to 2020 represents the predicted incident
totals for those years. The blue lines above and below the predicted values show the 95 percent
confidence interval (meaning that statistically there is a 95 percent chance that the actual incident
total will fall within that range).
Figure 7 and Table 6 that follows show the total incident projections and the projections
for various incident types.
Figure 7: Actual (solid black) and Predicted (red) Emergency Incidents by Type
Total Incidents
12000
10000
=
8000
6000
4000
-
2000
o
2000 2005 2010 2015 2020
Time
~l---i i
E
2000 2005 2010 2015 2020 2000 2005 2010 2015 2020 2000 2005 2010 2015 2020
2000 2005 2010 2015 2020 2000 2005 2010 2015 2020 2000 2005 2010 2015 2020
2000 2005 2010 2015 2020 2000 2005 2010 2015 2020 2000 2005 2010 2015 2020
2000 2005 2010 2015 2020 2000 2005 2010 2015 2020 2000 2005 2010 2015
,I
2020
From the analysis it was determined that total incidents appear to be very slightly
increasing. Over the last decade, total incidents have increased from 6207 in 2000 to 7366 in
2009, a 19 percent increase. Over the same time period, EMS incidents increased from 2742 to
4070, a 48 percent increase! We also note that the total number of fires has remained relatively
unchanged over the last decade, staying between 201 and 275 fires per year. Should these trends
continue, we would expect 9538 total incidents, 5842 EMS incidents, and 172 fires in 2020.
Again, the calculation is predicated on a multi-linear regression that includes population
predictions. For the city it should consider the blue line (confidence interval ranges) rather than
the actual prediction. The confidence interval ranges do show that statistically it can be expected
that total incidents will increase slightly over the next ten years, mostly due to increased EMS
incidents. The number of fires is expected to remain relatively unchanged.
Although there are certainly other incident types that show increases or decreases over
the last decade (e.g. Alarms and Hazardous Condition), do not read too much into these trends.
These graphs are based on information entered by firefighters after returning from an incident.
Because different people may code the same incident in different ways (there are hundreds of
different incident type codes), there will be discrepancies in the data.
Typically fires and EMS incidents are fairly straightforward to code and, as a result, we
place much more value in the reliability of those trends; they also happen to be two of the more
important incident types when it comes to fire department workload. Total incident counts
should also be fairly reliable because these counts do not depend on how the incident type was
coded.
Demand by Planning Neighborhoods – The following graph and table show emergency
services demand by planning area. The length of each bar represents the total number of
incidents that occurred in that planning area from 2007 to 2009. Within that bar, the different
incident types are represented by different colors depicted in the legend. Please note that these
values are not normalized. Planning areas that have particularly large land areas may have high
emergency services demand, yet have relatively low emergency services demand per capita or
per square mile.
Figure 8: Total Emergency Service Demand by Planning Areas, 2007-2009
Can- Good
EMS Alarm celled Intent Service MVA Other Fire Hazard
Fairmeadow 620 179 74 95 32 32 53 42 32
Leland Manor / Garland 534 175 103 75 75 48 24 39 24
Midtown / Midtown West 592 124 80 63 63 31 42 33 25
Barron Park 566 133 62 76 63 31 31 37 23
Old Palo Alto 528 117 92 76 62 45 28 18 18
Community Center 495 103 135 58 45 42 22 22 26
College Terrace 551 86 57 69 52 46 6 11 34
Palo Verde 528 91 70 56 53 22 22 31 16
Southgate 482 112 112 32 16 16 32 0 48
Duveneck / St Francis 451 93 88 44 59 34 25 20 22
Crescent Park 423 127 83 51 40 24 22 35 28
Professorville 403 113 38 38 101 50 13 0 76
Green Acres 447 122 68 48 48 29 20 9 14
Research Park 272 131 57 31 18 23 17 17 14
Stanford 159 78 27 19 13 10 12 10 6
Bayshore Area 26 7 5 3 1 3 2 2 2
Palo Alto Foothills 12 4 5 1 1 2 1 1 0
Fire Type Trends – There has been very little change in the number of fires from 2000-
2009. Vehicle and confined structure fire categories do have some up and down, but for such a
specific category that there are less than 60 incidents per year, it is expected that there be some
annual variability. Although PAFD should always work towards reducing fires through
prevention efforts, the lack of any significant increase in fires while population has very slowly
increased is good.
It should be noted that in the following figure, structure are broken out into several,
somewhat unusual classifications. These are not our choice of classification, but rather how the
US Fire Administration NFIRS specifications code fire types. The following is brief explanation
of these different structure fire types:
Structure Fire (confined) – This code is used to describe a fire that was inside a
structure, but confined to something like a cooking pot.
Structure Fire (non-building) – This code is used to describe a fire that occurred in
a non-building structure such as a bridge or pier.
Structure Fire (portable) – This code is used to describe a fire that occurred in a
portable building such as a mobile home or portable classroom trailer.
Structure Fire (other) – This code is used to describe structure fires that cannot be
classified into one of the other categories.
Structure Fire – The NFIRS coding includes catchall codes that describe an incident
type, but do not include details. This code is the catchall code for structure fires.
Figure 9 shows actual and forecasted fire incidents from 2000 to 2020. The red line in
this graph shows the predicted number of fires using the same multi-linear regression prediction
methodology as earlier. Blue lines above and below the prediction show the 95 percent
confidence interval. The black line from 2000 to 2009 shows the actual fire incident counts.
Total Fires
300
200
100
o
2000 2005 2010 2015 2020
Time
2000 200S 2010 2015 2020 2000 200S 2010 2015 2020 2000 200S 2010 2015 2020
ij l~!:--
~E, ,. I ij j::E;
• I :. 1~• ,I
2000 200S 2010 2015 2020 2000 200S 2010 2015 2020 2000 200S 2010 2015 2020
2000 200S 2010 2015 2020 2000 200S 2010 2015 2020
Jurisdiction-Wide Fire Losses – One of the best indicators of fire risk is actual data
collected from fires over multiple years, preferably at least five. The following table shows total
fire deaths, injuries, and property loss (defined as both the property and contents).
Table 9: Total Fire Loss, 2000-2009
Average
Loss ($)
Year Total Fires Dollar Loss Injuries Fatalities Per Fire
2000 195 $2,168,650 1 1 $11,121
2001 237 $1,482,510 0 0 $6,255
2002 236 $428,620 0 0 $1,816
2003 242 $78,190 0 0 $323
2004 187 $46,315 1 0 $248
2005 199 $439,470 2 0 $2,208
2006 181 $1,408,040 1 1 $7,779
2007 202 $6,992,390 0 2 $34,616
2008 176 $1,687,128 0 0 $9,586
2009 209 $1,297,305 0 1 $6,207
(average) 206 $1,602,862 .5 .5 $7,781
The data reflects all fires, including vehicle fires and outside fires. It appears that, on
average, there are around 200 fires per year with a fire injury or fire death occurring every couple
years. Annual dollar loss due to fire varies from $46,315 in 2004 to $6,992,390 in 2009; on
average there was about $1,600,000 of fire damage annually. It should be noted that the total
dollar loss for fires in 2003 and 2004 seem too low to be believable. The fire department should
recheck its data, especially for these years. Clearly though, if the loss data is correct Palo Alto
has a very low average dollar loss per fire incident. This is good and probably attributable to
factors such as the, sprinklered and fairly modern building stock, as well as the education level of
residents and those who work in the city. Rapid response by competent firefighters also should
make a difference (the loss saved by fire operations can be measured, but few do—we have
developed an approach for estimating losses and lives saved by fire operations, it this is of future
interest.)
The following table compares Palo Alto fire loss data to regional and national averages.
For a relatively small jurisdiction such as Palo Alto, fire loss data can be easily skewed. Any
slight under-or over-reporting of deaths, injuries, or property loss could have a huge impact on
the validity of the results. So comparisons should be viewed over time, to be based on more data.
Comparisons also should be taken with a grain of salt unless it is known that the data in different
jurisdictions is collected the same way.
Table 10: Per Capita Fire Loss and Comparison Statistics, 2000-2009
Palo Alto has fewer fires and fire deaths per capita than the averages for the US, the west
coast, and communities of similar size. Palo Alto also has a much lower average dollar loss than
the US and communities of similar size, but its dollar loss is just slightly higher than the region
and population. Fire injuries are significantly lower than national, regional and community size
averages. Again, that Palo Alto is doing so well is related to community demographics in part,
such as much higher levels of education compared to other, similar-size communities.
The median home value in Palo Alto makes a true comparison of fire loss difficult. As
compared to Palo Alto, where the median value of a residential property in 2010 was $1.41M,
the median value for residential homes in the rest of the Nation was approximately $167,000. 56
For Palo Alto, even a small fire inside a structure can result in a substantial fire loss and thus
make the overall fire situation appear high when it is not. Whereas a room-and-contents fire in
the typical U.S. community might exact a loss of $10-$30K, the same fire in Palo Alto might
easily result in a $150-$250K loss, or more. The point is, even in those years where fire losses
exceeded $1.0M, total losses are equivalent to one residential structure, not including its
contents.
Geospatial Location and Severity of Fire Incidents – To help evaluate the
appropriateness of fire unit locations, it helps to understand where there are fire ―
hotspots‖
(higher numbers and density of calls) within the city. It is also good to know where the more
serious structure fires (those that involve deaths or large fire losses) are occurring. Figure 10
shows both the fire density per square mile for all fire types (including vehicle and outside fires)
and the location and severity of structure fires shown by red dots that are scaled based on the
amount of fire loss. Additionally, we mapped out the location of all the civilian fire deaths shown
with green crosses. This map was compiled from PAFD incident data from 2007 to 2009.
Please note that one of the issues we ran into with this project is that the dispatch does not
hard-code coordinates for each incident, so we had to geocode the addresses that were provided
in the fire department incident reports. Like all other fire departments, these addresses can be
quite poor because of misspellings and omissions. For this study we were able to achieve a 97
percent match rate of incident geocodes by reducing the number of years of data analyzed.
5
http://www.1siliconvalley.com/palo-alto-housing-market-%e2%80%93-fourth-quarter-2010-update/
6
http://money.cnn.com/2010/05/11/real_estate/latest_home_prices/index.htm
Figure 10: Major Incident Locations: Dollar Loss and Fire Deaths, 2007-2009
'""
!
I
.--
'\
, --.......
---- ·=--..
S tructu... F"
" '" . 2007·2009
-""', ,..-
~ --"""'-'
-c. 0.-. ... 0 -
T........ l ... .
• _ _ $>',000
• "',""-"",000
• "",000_1'''',000
• I"OOO, · IOOO.OOIJ
• ", _ _ 1",",000
,.
Fire Risk by Planning Neighborhoods – Fire Risk is the product of fire probability and
fire consequence. High risk, therefore, can result from either a large number of small fires, or a
small number of large fires. The table below shows the average annual number of fires in each of
the planning areas and the number of fires per 100 square miles. Some of the planning areas were
larger than others, so it made sense to normalize the numbers in a way that allow accurate
comparison. The fire ― probability‖ classifications (low, medium, and high) are based on which
quartile the fires ―
per 100 square miles‖ value fell into. The top 25 percent where classified as
high fire probability, the middle 50 percent classified as medium fire probability, and the lowest
25 percent classified as low fire probability. Charleston Terrace, the Commercial Area,
Downtown North, Evergreen Park, Fairmeadow, University Park, and Ventura have the highest
probability of fire.
The same table also assesses fire ―severity‖ using financial fire loss as a metric. Again,
the statistics were normalized by land area (fire loss per square miles) and ― severity‖ classified
using the quartiles. Using this methodology, it was determined that Byron Park, the Commercial
Area, Community Center, Downtown North, Leland Manor/Garland, Palo Verde, Research Park,
and University Park had the most severe fire loss.
We have already discussed how this table shows us the probability of fire and the actual
outcome (consequence/severity) for each of the different planning areas. Although we used a
fairly simplistic quartile method for classifying the planning areas, it was a quick way of judging
probability and consequence of fire in different geographic areas of the city. We use these to
inputs to make a final fire risk classification shown in the far right column of the same table.
Generally speaking, a high/high or high/medium combination was classified as high risk. A
low/low combination was classified as low risk. The remaining planning areas were classified as
medium risk. The Commercial Area, Downtown North, and University Park appear to have the
highest overall fire risk.
Table 11: Overall Fire Risk Classification by Planning Area, 2007-2009
Total Fires (per sq mile) Probability Fire Loss (per sq mile) Severity Risk Level
Barron Park 23 37 Medium $325,630 $527,680 High Medium
Bayshore Area 9 2 Low $4,615 $1,057 Low Low
Charleston Meadows 11 35 Medium $123,601 $394,537 Medium Medium
Charleston Terrace 30 45 High $24,710 $37,423 Medium Medium
College Terrace 2 11 Low $400 $2,298 Low Low
Commercial Area 62 93 High $498,125 $744,217 High High
Community Center 7 22 Medium $536,510 $1,723,948 High Medium
Crescent Park 19 35 Medium $900 $1,653 Low Low
Downtown North 11 58 High $80,050 $422,860 High High
Duveneck / St Francis 8 20 Medium $62,100 $152,234 Medium Medium
Evergreen Park 16 75 High $31,250 $146,850 Medium Medium
Total Fires (per sq mile) Probability Fire Loss (per sq mile) Severity Risk Level
Fairmeadow 4 42 High $1,050 $11,032 Low Medium
Green Acres 4 9 Low $0 $0 Low Low
Greenmeadow 11 28 Medium $1,202 $3,035 Low Low
Leland Manor / Garland 13 39 Medium $306,000 $923,944 High Medium
Midtown / Midtown West 39 33 Medium $26,790 $22,913 Low Low
Old Palo Alto 11 18 Medium $14,625 $23,697 Medium Medium
Palo Alto Foothills 8 1 Low $72,200 $6,561 Low Low
Palo Verde 24 31 Medium $515,515 $672,558 High Medium
Professorville 0 0 Low $30,000 $377,948 Medium Low
Research Park 33 17 Medium $200,244,770 $105,586,541 High Medium
Southgate 0 0 Low $0 $0 Low Low
Stanford 111 10 Low $321,796 $29,037 Medium Low
University Park 73 243 High $6,573,183 $21,900,755 High High
University South 4 20 Medium $30,000 $148,560 Medium Medium
Ventura 14 40 High $55,371 $157,397 Medium Medium
Figure 11: Correlation between Population and EMS Calls for Service Study, 1980-2004
With this correlation established, it is possible to look at a population density map of Palo
Alto and predict which areas will have the highest EMS demand per capita. The following
population density map (Figure 12) was compiled by ESRI using United States Census data.
EMS Incident Density – Based on the Palo Alto population density map, a high number
of the EMS incidents should be expected in the downtown area and on the University campus
where there are the largest concentrations of residential population. Besides having large
residential population densities, these two areas also attract a large number of commuters during
business hours. We would also expect to see EMS hotspots in the eastern portion of Ventura
neighborhood, the northwestern portion of Midtown, and the eastern portion of the Charleston
Meadows neighborhood where there are smaller areas of high population density. The
actual number of EMS incidents by planning area is listed in the next section.
We see in the EMS incident density map (Figure 13) that the downtown areas have the
highest number of EMS incidents and there are small hotspots along Alma Street. Despite a high
population density, the Stanford Campus does not have high EMS demand probably because
young students with limited chronic medical problems comprise most of the residential
population in this area.
1""•.....,
•
I
-- I
/~ ..
Each planning area is assigned an EMS risk classification below based on the number of
EMS calls per 100 square miles. Based on our risk evaluation, it was found that Charleston
Meadows, the Commercial Area, Downtown North, Evergreen Park, and Greenmeadow have the
highest EMS risk.
Table 12: EMS Risk Factors and Classification, 2007-2009
Target Hazards
For this project we also assessed as part of the analysis the location of target hazards
provided to us by the fire department. These are h of these target hazards are shown in Figure 14.
Although we do not provide an explicit review of these target hazards due to the scope of this
study, we included this map because it is something that we do refer to when making station
location recommendations in the next chapter.
A,flO_
""0oAllo
f '~m.
c:~ ... c:o .......~
C Stanford Boundary
"
[-=--=--=-1 Palo Alto Boundary
Target Hazards -"'
• Commercial
•o Population Cente r
Transportation & Ene rgy
The major steps for a deployment analysis include a risk assessment (discussed in the
previous chapter), working with the public and local government officials to determine response
time goals for the community as a whole or by individual planning areas, and measuring current
and potential performance against selected goals. The Center for Public Safety Excellence
(CPSE) publishes an excellent reference that can be used by communities to understand the
process and determine the choices available to them. Generally referred to as a ‖standard of
cover‖ analysis, we used the CPSE methodology in the analysis of response time and service
reliability.7
Strategic Goals
Establishing goals against which to measure response effectiveness rests with PAFD
leadership. The following steps should be considered:
Discuss with community leaders and local government officials the desired
performance measures for fire and EMS response times. Very few citizens and
elected officials understand the concept of response time as it relates to emergency
services. Educating them on the available choices and trade-offs are necessary parts
of the planning process. For example, it is useful to note to them that rapid response
may prevent a food on the stove fire from becoming a house fire; a cardiac arrest from
becoming a fatality; and a gunshot wound from becoming a murder.
7
Creating and Evaluating Standards of Response Coverage for Fire Departments, Commission on Fire
Accreditation International, Inc., Chantilly, VA 20105.
Implement a method for tracking unit reliability. The CPSE Standards of Cover
manual places great importance on considering deployment changes when a unit falls
below 80 percent performance reliability. Reliability in this case means whether the
unit located at the closest fire station is available to answer the call.
Prepare a monthly response time report and distribute the report department wide.
(TriData could help set up such a report).
Conduct an annual performance review for the entire system and each planning area.
The report should summarize the systems overall performance, note changes to
specific planning areas, and recommend deployment modifications, near and long-
term.
NFPA 1710 provides generally accepted response time standards for career fire
departments, though there is no single set of nationally accepted response time standards. Many
communities choose to develop their own response time goals in light of what is currently
achieved versus what it would take to improve them. There have been a few attempts to measure
the incremental value of a minute faster response time for fires and EMS calls, but there is no
definitive study of the incremental benefit. Faster is better, but it is unclear how much better in
terms of dollars or lived saved.
Most fire departments use the NFPA 1710 standard as a goal, not as a prescriptive
requirement. Few departments are currently meeting or exceeding NFPA 1710, especially with
respect to travel time (which is the hardest to improve). Palo Alto travel times are only about a
minute over the NFPA 1710 recommendations, and this is likely the result of difficult traffic
patterns and an abundance of road blocks rather than a lack of fire stations. There are also issues
with overall response times, partly because of increased travel times, but also related to
questionable call processing times and poor turnout times. Data collection techniques and data
reliability is also an issue as these time segments seem to be inconsistent.
First, response times are subject to a variety of measurement errors and only measure one
aspect of overall system performance. For example, response times are distorted when units
report their arrival on scene either early or late. Second, response times are frequently not
comparable across fire-rescue systems because of the differing manners in which they are
calculated. Not all departments track vertical response times (that is, the time from arrival on
scene to patient contact), so their total response times likely would be lower than the total
response times of the few departments that do track them.
Many fire/EMS departments report average response times while others report fractile
response times. Average response times have been increasingly less used by the emergency
service industry because small numbers of very short or long responses are often recorded in
error. Also, one or a few long response time s(or errors) can distort the results. The public is
interested in how fast a system responds to most calls, which is better reflected in fractiles rather
than averages. More and more departments are adopting the 90th percentile for reporting
response times (mostly due to NFPA 1710‘s use of this measure). However, meeting the 90th
percentile goal is not always the most efficient means for delivering emergency services. A
fractile response time of x minutes at the 90th percentile means that at least one unit responds in
x minutes, 90 percent of the time. The remainder beyond the compliance fractile (10 percent in
this case) is the operational tolerance for the system, meaning the system is designed with the
understanding that 10 percent of the calls may have response times that exceed the target.
Although it is possible to design a system that may ensure rapid response close to 100 percent of
the time, it is generally not cost-effective.
Most departments, including PAFD, do not record the vertical response time component.
However, given the high number of areas in Palo Alto that require additional time to reach a
patient (e.g. Stanford buildings, high-rise buildings, et cetera), it would be a good idea to start
measuring vertical response time, especially for EMS calls. Vertical response time affects the
realistic expectations for cardiac resuscitation outcomes, and should be considered in response
strategy, such as deployment of AEDs, training neighbors in CPR, and deploying more units in
high-rise areas. A longer interval of collapse to care time lessens the likelihood of successful
resuscitation. Travel times of four minutes to a bedroom community are likely to yield different
patient outcomes than to urban, high-rise communities, as ‖with patient‖ times increase the time
to care and decreases the probability of successful resuscitation.
Another aspect of measuring the time to be ‖with patient‖ or ‖at fire‖ is better quality
oversight. For incidents such as cardiac arrest or fire showing, crews could report additional
intervals such as ‖AED placed‖ or ― water flowing,‖ allowing PAFD to measure some aspects of
skill quality. For AED placement times, technology and clock synchronization may be used in
lieu of extra radio transmissions. Reportedly the fire department is already collecting the time
until the application of an automatic external defibrillator (AED). However, the data is not
routinely analyzed.
While the speed of response is not directly indicative of service quality, it does affect the
number of lives saved and the value of property losses averted when an emergency occurs. This
means that while arriving in three or four minutes every time does not guarantee everyone will
live and there will be less damage, more people can be helped or the fire can be extinguished
before it becomes severe when emergency personnel arrive in 5 minutes rather than 10 or 20.
Fire spreads quickly after ignition and the faster it is found and extinguished, the better the
results; similar to someone suffering from life threatening symptoms, the probability of survival
increases the quicker the patient is treated. There are also real times limits as to how long one
can go without breathing or when bleeding profusely.
Even with general observations and response-time analyses, current statistical models
cannot realistically assess nor predict the quality of fire services in terms of lives saved and
property losses averted. In place of true measures of fire rescue service outcome, response time is
often used as a proxy measure.
According to multiple studies, extension of the fire beyond the room of origin typically
begins before 6 minutes after ignition, and flashover of the room of origin occurs within 10
minutes of ignition. (Flashover is the simultaneous ignition of all flammable material in an
enclosed area.) In some modern rooms with low ceiling and plastics, flashover can occur in two
to four minutes, according to studies by the National Institute of Standards and Technology.
Figure 16 shows the fire propagation curve, which shows the effect of time and temperature rise
of a free-burning fire on the destruction of property.
Figure 16: Fire Propagation Curve
The fire propagation curve above is based on a typical unsprinklered room. Palo Alto fire
and city officials should keep in mind that the public investment in sprinkler systems may affect
what are reasonable response time standards. Building design and sprinkler systems can be used
to hold fires in check for longer periods of time. When an entire area is sprinklered, response
times for fire incidents can be increased. However, unless a whole area is sprinklered, there are
still many unsprinklered buildings that need rapid response. Later in this chapter, we discuss
setting both city-wide response time and planning area response time goals. The percentage of
buildings sprinklered should be considered when setting response time goals for fire incidents.
Call Processing – Call processing time in theory should include both call-taking and the
actual dispatch of the equipment. Call-taking is the time to get information from the caller and
enter it into the dispatch computer system. This is measured from the time the call is received to
the time the call is transferred to a dispatcher. Dispatch time begins when the call is transferred
from the call-taker to a dispatcher and continues until units are alerted to respond. The NFPA
1710 Standard recommends a 1:00 minute dispatch time.
PAFD dispatch times were found to be particularly low and it is unlikely that the times
we received for analysis reflect the total time that dispatch took . Although we were not able to
verify the cause, it appears that the dispatch times analyzed reflected only the dispatch portion of
the time segment and not the time it took to handle the call and record the information from the
caller. (In other words, the time should reflect call processing including dispatch.)
Table 14 and Figure 17 show the call processing and dispatch times by time of day and
incident type. The red line depicts the call volume by time of day. The line clearly shows the
highest volume of calls peaking between 8:00am and 6:00pm, when call volume begins to
decrease.
Table 14: 80th Percentile Call Processing Time by Incidents Type
Turnout (or Reaction) Time – Turnout time is the second segment in total response
time. It begins when personnel are alerted at the station for the emergency until the crew is
aboard the apparatus and ready to respond. If data collection is accurate, it can be calculated as
the difference between the en route time and the unit dispatch time. The goal for turnout is also
one minute, 90 percent of the time. However, this goal is sometimes difficult to achieve for most
departments and a more realistic goal is one minute for daytime calls and one and one-half
minutes for calls occurring during times when firefighters are not awake. Turnout times for non-
emergency calls also need to be eliminated from the data set since firefighters do not react as
quickly if the call is not an emergency.
Even when a more realistic goal of 90 seconds for turnout is considered, the PAFD is still
exceeding the goal considerably. This should be an area for concern and the reasons investigated.
Table 15 and Figure 18 show the turnout times by time of day and incident type. As
expected turnout times during night time hours were longer, but the difference is not as
pronounced as what is found with most other fire departments.
Table 15: 80th Percentile Turnout Time by Incidents Type
Findings of excessively long turnout times led to our having doubt as to the accuracy of
the data since the turnout time for firefighters in Palo Alto significantly exceeds those found in
most other communities. For this reason, the fire department should revisit the data and conduct
more study of turnout. It should r disaggregate the results by fire station (and shift) to determine
if there are problems at a particular station. Longer turnout times can sometimes be related to the
layout of a particular fire station more so than to a particular crew, but crews also need
encouragement to respond more quickly too. Fire stations in Palo Alto are not overly large, and
there are no architectural layout reasons that it should take so long for personnel to don their
turnout equipment and mount the unit once alerted.
Ultimately, the issue of long turnout times needs to be addressed because these times can
be improved more easily and with little expense than moving stations. Once the apparatus leaves
the fire station, response times are more difficult to improve since the speed is related to the type
of roads traveled, weather, and traffic.
Recommendation 16: Revisit the analysis of turnout times and take the necessary steps to
improve them where possible. Establish a performance goal for each time segment and assess
them monthly. Evaluate turnout times by fire station and shift.
Travel Time by Hour of the Day and Incident Type – Travel time starts when a unit
begins its response from the fire station until it arrives at the scene. Travel times are a function of
geography, road conditions, traffic/congestion, and the number of and location of fire stations
with respect to the location of actual calls. A travel-time goal of four minutes is suggested by
NFPA. Later in this section we discuss in more detail the location of fire stations and the
performance of the current layout of stations.
At the 80th percentile, travel times ranged from 4:30 for EMS calls to 5:22 for hazardous
condition calls. These times are all higher than the NFPA 1710 recommended 4:00 travel time.
As noted earlier, Palo Alto has some unique traffic patterns. There are not a lot of through-
streets, many barriers are in place around Stanford that restrict travel, and quick response times
through residential neighborhoods can be difficult because of the narrow streets that reduce
driving speed. We expected that daytime traffic would also impact travel times, but there was
little difference in travel times during different times of the day.
Although PAFD should clearly continue to try and reduce the travel time to emergencies,
it is not a realistic goal to reach the 4:00 NFPA goal 90 percent of the time. A better approach is
to improve call processing and turnout times to compensate for slightly longer travel times.
Table 16 and Figure 19 show the travel time for the first unit to arrive at an incident.
Table 16 depicts the travel time categorized by incident type. Figure 19 show the travel time by
hour of day with the call volume shown for reference.
Table 16: 80th Percentile Travel Time (First Unit) by Incidents Type
As can be seen in Table 16, when a goal of 80 percent is used, the response time for the
first unit to arrive is one minute faster (4:42 versus 5:42) for all calls.
Figure 19 shows the 80th percentile travel time by hour of the day for the first-arriving
unit. Call volume for the 24-hour period is also depicted.
Figure 19: 80th Percentile Travel Time (First Unit) by Hour of the Day
Total Response (Reflex) Time – As discussed earlier, total response time is measured
from the time the call is received by the dispatch center until a unit arrives on scene. Total
response time is, for most citizens, the most important time element and the measure they use to
evaluate the effectiveness of fire and EMS service.
A review of the total response time was conducted using NFPA 1710, which recommends
a six minute total response time. Again, this is based on one minute for call processing, one
minute for turnout, and four minutes for travel time.
PAFD is actually pretty close to this goal time at the 80th percentile level (about a minute
off for most incident types), but remember that the short call processing times may be an
indication that we are not looking at the total response time here; it may be necessary to add
about 30 seconds to these total response times to account for a missing call taking segment.
Problems with response time, when they are found, are usually with station locations, or
not having enough stations to cover the area. Such is not the case in Palo Alto where it appears
that a combination of long turnout times and a difficult travel environment are leading to long
response times. In our judgment response time issues are not the result of station location issues,
or having too few stations.
Table 17 shows the total response time for the first unit to arrive at an emergency. Figure
20 shows total response time by time of day. As can be seen there is little difference in total
response time between daylight and night time hours, which is unusual.
Table 17: 80th Percentile Total Reflex Time (First Unit) by Incidents Type
Figure 20: 80th Percentile Total Reflex Time (First Unit) by Hour of the Day
Table 18 shows the travel time, fire department reflex time, and total response (reflex)
time for each of the fire department first-due areas. Travel times are not great, but okay. Incidents
in Station 2‘s first-due area have a travel time of 4:04, likely the result of more open roads with
fewer traffic lights and stop signs in their area of response. Travel times for the remaining first-
due areas are all in the 4:30 to 5:00 minute range with Station 3 having the longest travel time of
5:18.
Station 1, 2, 4, and 5‘s first-due areas all have total response times that are within a
minute of meeting NFPA 1710 at the 80th percentile level. Also included in the table are fire
department reflex times (FD Reflex) which are the times from unit dispatch to unit arrival.
Table 18: 80th Percentile Total Response Times by First Due Area
The first look at incident data showed that downtown Palo Alto had a much larger
number of calls and more resources to handle the call volume. If the city-wide performance
standard was to achieve a certain response time 90 percent of the time and the downtown
accounted for the largest portion of calls, performance in other areas of the system could well be
substandard. For this reason a higher level of analysis was needed as opposed to a smaller
community where call volumes and incident types are often homogenous across the entire
system.
In the previous chapter, we developed planning areas based on the historical
neighborhoods to conduct the risk and demand analysis. We used this approach because planning
areas run along neighborhood boundaries and both risk and demand are influenced by
neighborhood characteristics. Also, using planning areas helps non-fire department officials and
the public better understand the results (because people are typically much more familiar with
neighborhood boundaries than fire districts or first-due boundaries).
The analysis of fire station and apparatus locations is based on overall deployment
performance (workload, travel times, and performance reliability). Planning areas in Palo Alto
are sometimes protected from several fire stations and the areas covered by individual fire
stations do not match the city‘s planning area boundaries. If deployment performance was
measured using planning areas, it would be difficult to attribute problems to a particular fire
station. For that reason, our analysis looks at workload and deployment performance by fire
stations and the first-due areas they covered.
In addition to performance and travel times, we also analyzed the workload for each fire
station and each unit. This determined whether particular units could handle more calls without
compromising the overall performance of the area.
Performance Goals – As part of the planning process, PAFD will have to decide the
performance goals it considers right for the city overall and for each planning area. It will also
need to find a reasonable response time and a reliability goal for each planning area (and each
incident type).
Setting performance goals by planning area is one way to do this because risk factors can
be matched with travel time and reliability goals. Then, if analysis shows a problem with
response times (or reliability), further analysis can be done to determine the contributing factors.
Another way is to create demand zones across the city. However, the planning areas (essentially
historic neighborhoods) appear to be a good framework. Ultimately, fire leaders need to
understand why the desired goals are not being achieved in a particular area and the response
situation for each fire station protecting that area before it makes changes to the system.
As stated earlier, the NFPA 1710 response time standard is based on typical fire growth
rates and patient outcomes, primarily those involving cardiac arrest. The recommended time for
the first unit to arrive under the standard for both fire and EMS incidents is six minutes (four
minute travel time plus two minutes for call processing and turnout time). The time is based on
research showing that a structure fire begins to grow exponentially after six minutes and
individuals in cardiac arrest need defibrillation within six-minutes.
The problem with using standards ― carte-blanche‖ is that they are sometimes overkill for
the particular situation. For example, an area with a very young population might be okay with
an eight-minute medical response time since the more serious and time-sensitive EMS calls
occur less frequently. Likewise, an area with a large percentage of sprinklered buildings might
not require as fast of a response as those in unprotected buildings. Planning areas where a
majority of structures are equipped with sprinklers, can have a lower performance goal (80
percent or even 75) applied as the acceptable goal.
Appropriate performance levels are very much based on the characteristics of individual
planning areas. Fire department personnel are very good at determining appropriate response
time and reliability goals. For its part, PAFD should have its strategic planning team, and others
within the fire department familiar with the various planning areas, recommend the response time
goals for each of the city‘s planning areas.
Recommendation 17: Use NFPA 1710 (and other standards) to develop performance goals,
but consider each planning area on the merits of its particular situation.
Assessing Deployment Performance – Assessing fire department deployment is a
difficult task because of the many factors that affect performance. A simplistic way of
determining fire station locations would be to use a GIS program to map out four minute
coverage areas to make sure there are no coverage gaps. This method focuses entirely on the
location of the fire station and would work well if the fire department only answered one call at a
time. The problem of this approach for a city like Palo Alto is that some stations are busier than
others and concurrent calls are common, especially during weekdays.
Using the performance analysis method, we determined whether response time goals
were being met regardless of which fire station or unit may have responded to the call. This
permitted us to consider unit workloads and whether particular fire units could handle more calls
before performance was affected.
Standard of Cover Evaluation – For this study we used the Center for Public Safety
Excellence (CPSE) Standard of Cover process to evaluate the overall performance in Palo Alto.
The premise of this assessment method is that for each analysis area, there is a trade-off between
unit availability and performance. Generally speaking, as the correct unit for a particular area
becomes less available (due to other calls, training, etc.), performance for that area should
decrease because outside responding units from other stations have further distances to travel.
For this analysis we use two metrics:
Unit Availability - the percentage of incidents where a unit from the correct station
was available to handle the call and did so.
Performance Level - the percentage of incidents where the travel time was at or below
(faster) than the recommended goal.
In order to run this analysis, it is necessary to set a travel goal time and goal by which to
measure the performance level. In most cases, we use the NFPA recommended time of 4:00, but
in Palo Alto it appears that the road network makes it very difficult to achieve a 4:00 travel time.
As a result we have set a 5:00 travel goal for this performance level vs. unit availability analysis.
Figure 21 shows an example performance level vs. unit availability analysis.
In the example above the dashed black line in the performance level vs. unit availability
analysis marks the 80 percent performance threshold. If the dot for a particular station falls
below that dashed line, response travel time performance is being affected and changes may need
to be considered. In addition to showing whether performance is being achieved, the
performance graph can also show workload sensitivity or fire station location problems.
The slope of the line assesses workload sensitivity. A nearly horizontal line indicates that
unit availability has little impact on performance levels, meaning that the area in question has
low workload sensitivity. Fire stations with low workload sensitivity typically have many nearby
fire stations that can cover their calls if they are unavailable. A line that slopes strongly
downwards indicates that performance levels are heavily dependent on unit availability. These
areas are very workload sensitive because second-due units are generally unable to achieve travel
time goals when the first-due unit is unavailable.
Station location problems can also be identified by this analysis. If this part of the line is
under the 80 percent performance level threshold it means that, even when the correct unit
responds, travel time goals are not being met 20 percent of the time. This means that the station
is not well located to reach all parts of its first-due area. This problem can be fixed by relocating
the station, building a new station, or (perhaps the best solution) re-evaluating first-due
boundaries to make sure that the most appropriate station is responding.
The CPSE recommends that as the actual performance level falls below 80 percent,
measures should be taken to address the issue. This 80 percent level does not match up with the
90 percent compliance that NFPA 1710 recommends. However, the 90 percent reliability goal is
an appropriate threshold for city-wide performance; it is a very difficult to meet this goal in every
area of the city. In fact, the very low call volumes in some areas make it unrealistic that the city
would want to achieve such a standard since it would be inefficient to provide enough fire
stations to meet the standard all of the time.
All things considered, 80 percent compliance is an appropriate goal. That means that
first-due areas where less than 80 percent of calls are reached within the prescribed travel time
(which we suggest should be five minutes for Palo Alto) should be considered problem areas.
Palo Alto Performance Analysis – As we have already discussed, Palo Alto has a road
network with lots of stop signs and barriers that makes it difficult to achieve the typical 4:00
NFPA 1710 travel time goal. Although, ultimately the travel time goal determination rests on the
public officials and fire department decision makers, we have chosen to assess performance and
reliability in the context of a more realistic travel time goal of five minutes.
Figure 22 and Table 19 show the performance vs. reliability assuming a 5:00 travel time.
Figure 22: Performance Level vs. Unit Availability for 5:00 Travel Time Goal
Table 19: Performance Level vs. Unit Availability for 5:00 Travel Time Goal
Assuming that 5:00 is the desired travel time goal for Palo Alto, Stations 5 and 2 have a
fair amount of additional capacity (meaning that those stations can handle additional workload
without affecting their performance level). Stations 1, 4, 6, and 7 are currently right at capacity.
Station 3 appears to have a station location problem because even when unit availability is 100
percent, this station is just barely achieving an 80 percent performance level.
We also note that Stations 1, 5, and 7 have fairly horizontal lines, which means that when
those stations do not have a unit available, other stations are able to cover their calls within the
performance standard. Stations 2, 3, 4, and 6 have a much higher negative slope as unit
availability decreases. This means that other stations cannot cover calls in these areas within the
performance standard if the correct unit/station is unavailable to respond.
Assuming that all data provided to us is accurate, the following points can be derived
from the reliability and performance analysis.
Stations 4 and 6 are currently performing right at the 5:00 travel time goal. These two
stations have a decreasing performance level as unit availability decreases. As a
result, we would not want to add any additional workload to these stations and
recommend that Palo Alto keep a close eye on their performance if workload does
increase.
Stations 7 and 1 are currently performing right at the 5:00 travel goal time, but
because their performance is not as workload sensitive, these stations could
potentially handle additional workload.
Station 8 was not evaluated because the number of calls was so low.
Workload Analysis – For this analysis we also looked at the call volume and workload
for each fire station and unit. Workload is defined as the total time (hours) that a unit is actually
engaged in providing service. EMS rescues that transport patients typically have higher
workloads than do fire engines.
As explained in previous section, these factors affect performance and reliability for the
reason that higher workloads for one unit reduce the availability of that unit for another call such
that the incident must be handled by another unit nearby. For example, a fire station with a high
workload (such as Fire Station 1) might continue to meet its goals because other nearby stations
can also cover the area adequately. Other areas however, (Fire Station 4) are more susceptible to
workload increases because of the larger distance between stations.
Table 20 shows the run totals by station and unit for 2009. All analysis in this section is
based on 2009 CAD data received from the emergency dispatch center. None of the units except
Medic 2 at Station 2 is over 3000 calls, which is roughly the benchmark for being ― busy‖ and
starting to need relief.
Table 20: Responses by Station and Unit
We can see from this table that station 2 was the busiest with 5,095 runs followed by
station 1 with 3,610 runs and station 6 with 2,156 runs. A lot of these runs are, however,
ambulance runs that do not necessarily occur in that station‘s first-due area. In the following
graphs and tables we show the workload (in unit hours) both with and without ambulances. We
will defer discussion of the workload to the last section of this chapter where we look at all the
factors affecting the different stations and make deployment recommendations.
Figure 24 and Table 22 show the workload and call types for each station.
Figure 24: Station Workload (Unit Hours)
BLS ST1 ST2 ST3 ST4 ST5 ST6 ST7 ST8 (all)
EMS 666.9 1719.6 2398.5 429.0 277.1 241.3 325.2 4.4 6.8 6068.8
Alarm 0.0 59.5 89.8 32.5 36.4 43.1 146.8 26.6 1.0 435.6
MVA 9.5 55.3 225.2 47.7 26.3 14.1 34.7 2.2 2.5 417.4
Hazard 1.3 27.6 99.0 27.9 26.4 20.5 48.7 5.3 0.0 256.9
Public Service 1.1 41.0 38.1 40.5 34.5 14.6 25.2 0.2 0.3 195.4
Canceled 1.4 29.0 48.8 22.8 10.3 7.5 25.8 5.0 1.2 152.1
Structure Fire 0.0 28.4 50.8 14.1 6.5 2.9 45.6 0.6 0.0 148.9
Good Intent 4.8 28.4 46.8 17.9 7.6 17.0 16.3 0.9 0.3 139.9
Outside Fire 0.0 13.2 21.1 8.9 16.3 10.4 24.8 1.1 5.0 100.7
(all) 685.0 2002.0 3018.1 641.2 441.3 371.4 693.2 46.4 17.0 7915.7
Figure 25 shows the workload by station for the various call types but without the ALS
rescue units included. This is important because its shows the workload for fire units (engines,
ladder truck and rescue).
Figure 25: Station Workload (Unit Hours) without Ambulances
As stated at the outset of this chapter, the current configuration of fire stations provides
good coverage. With the exception of the Bayshore area (golf course and airport) and the
foothills, all parts of the city have a fire station with six minutes drive time and most of the city is
within four minutes of a fire station.
Figure 27 shows the theoretical travel time from each of the city‘s 8 fire stations. Areas in
dark green can theoretically be reached in four minutes and areas in light green can be reached in
six minutes. Typically we discuss any coverage gaps that are found in the theoretical drive time
analysis, but Palo Alto does not appear to have any major coverage gaps. The southwest corner
of the Stanford Campus has a little area that is not, but this is minor issue.
Figure 27: Four-and Six-Minute Travel Time Analysis
Coverage Overlap – Simply showing that there is adequate coverage from all the fire
stations does not take into account that more than one incident often occurs at the same time in a
first due area. Some coverage overlap is good in the areas of highest demand. Figure 28 shows
how many stations can reach each part of the city within four minutes. Thinking back to the plots
of fire and EMS incident density in the previous section, we remember that the vast majority of
emergency services demand occurs in the downtown Palo Alto area just north and west of
Station 1. It is this area that we would want to have coverage from multiple stations; this is, in
fact, the situation. Stations 1, 3, and 6 can all reach the downtown area within 4:00 travel time.
There is also an area between stations 2, 5, and 4 that have a high amount of station coverage
overlap where the risk and demand level does not justify it.
Figure 28: Current 4-Minute Station Overlap
Below is a station-by-station review of our station location analysis and some possible
deployment improvements or efficiencies. We reviewed workload, response time performance,
unit/station reliability, and theoretical GIS drive times together in forming conclusions.
Fire Station 1 – Station 1 houses Engine 1, Medic 1 and Medic 3, the busiest engine and
second busiest ambulance. This is the second busiest station (first without ambulance workload)
and is located in the southwest portion of the downtown area. Medic 3 is cross-staffed by Engine
1. The performance vs. workload analysis showed that this station could take on additional
workload capacity, mostly because other stations (Stations 6 and 3) can still provide adequate
response times into the downtown area when units from Station 1 are not available. Because the
downtown area is one of the high-risk and high-demand areas, this is the type of deployment
performance one would want here.
At first look, this station appears to be poorly situated because of its close proximity to
the city boundary. It is often preferable that a station is not this close to the city boundaries so
that its response area in the city is maximized. At first glance it was believed that Station 1
would be better located slightly northeast of its current location where it would have a larger 6
minute travel reach, still be located in the high demand downtown area, and potentially eliminate
the need for Station 3. The problem with this scenario is that Station 1 serves the area along Sand
Hill Road to the southwest of the station that would no longer be reachable within 6 minutes if
the station is moved to the northeast. There have been several large developments along this road
and it is important that they have good coverage.
Fire Station 2 – Station 2 houses Engine 2, Rescue 2 and Medic 2. This station is in the
industrial research park Medic 2 is the busiest ambulance. The engine has an average workload
compared to other engines. The rescue unit is the city‘s only unit of its type and has a very low
workload. We found in the performance vs. reliability analysis that the station has additional
workload capacity; both because of high-availability and the fact that there is little performance
drop when units respond to its area from second-due stations (Stations 6, 5, 3, and 4). Reliability
at this station is quite high because there are three units at this station. Later in the chapter we
recommend merging Station 2 and Station 5.
Fire Station 3 – This station houses Engine 3, the third busiest engine and least busy
ambulance. The station is located in an area of high-end residential neighborhoods. The
performance vs. reliability analysis showed that this station is susceptible to high workloads and
is well below the travel time performance threshold. The station is fairly well inset from the city
boundary (to the north) and, because most of that area is residential neighborhoods, it may be
difficult to make good response times close to the border. Also, this-area is not close to any other
fire stations (Palo Alto or mutual aid), so when there is not a unit available at this station, it is
very difficult for other stations to make good response times into the northern and eastern
portions of this first-due area.
Recommendation 18: Take a closer look at what is causing response time performance issues
in Station 3’s first-due area. Consider adjusting the first-due areas for this station to increase
unit reliability.
This station also provides a good backup for Station 1 but, because of its poor
performance and workload sensitivity, this station should probably only back up Station 1 for
structure fires, cardiac arrests, and similarly serious, time-sensitive emergencies. An ambulance
should be kept at this station because it provides good response times into the high-demand
downtown area and it potentially decreases the fire engines workload keeping it available more
of the time (that would require only dispatching the engine on ALS calls).
Fire Station 4 – This station houses Engine 4, one of the least busy engines. The
performance vs. workload analysis showed response time performance right at the 80 percent
threshold. Any additional workload should be closely monitored because, as reliability falls off,
performance takes a hit. It is difficult for second-due units (from Station 5 and 2) to make the
performance goal into this area. The ambulance should be kept at this station to keep up the
reliability and PAFD should closely monitor this station for falling performance. From a
coverage perspective, this station appears to be in a fairly good location for the northeast corner
of Palo Alto.
Recommendation 19: Consider adding Mountain View Fire Department as second-due into
this area if unit-reliability falls and affects response time performance. When Station 4 does
not have a unit available, the closest stations are 3, 4, and 5. We are already trying to increase
Station 3 unit reliability, so we would want to avoid pulling engine 3 out of place. Station 2 and
5 would provide good second-due coverage, but if those stations are merged (as recommended
later in this chapter), it will be important to keep unit reliability up for that station. It appears that
both Mountain View Station 3 and Station 5 would have good response times into Palo Alto
Station 4‘s first-due area.
Fire Station 5 – Station 5 houses Engine 5, the least busy engine next to Engine 7. This
station has low workload, high reliability and high performance and is a good candidate for
consolidation. It also sits right on the Palo Alto border which is not good location for optimizing
response reach. Please see the later recommendation on merging Station 2 and Station 5.
Fire Station 6 – This station houses Engine 6 and Truck 6, the second busiest engine and
only truck. This station has a good location on the Stanford campus and provides good response
time coverage to the whole campus. This station does have some additional workload capacity,
particularly because the nearby stations (Stations 1, 3, 2 and 7) are able to provide adequate
response times into this area when units from Station 6 are not available. Because this area is
busy, this is the type of deployment performance we would want and expect here.
Fire Station 7 – It does not appear that Station 7 is needed in the greater Palo Alto fire
department deployment scheme. This station simply fulfills its contracted role as first responder
to all Stanford SLAC emergencies. Station 7 has incredibly low workload and is not in a good
position to respond to anything other than as a complement unit (which it cannot really do
because this station needs to be backfilled within 10 minutes.
Fire Station 8 – This station also has very low workload but protects some high-value,
high-risk properties. Recommendations for this station are discussed later.
Station 3 and Station 4 Merger – These two stations cover the largest combined first-
due area and their performance is strongly affected by their workload (because other units cannot
make good response times into their area when these stations are busy). Additionally, there was
no single station location that could provide good 4 and 6-minute coverage to their first-due
areas. We found no good merger opportunities for these two stations.
Station 1 and Station 3 Merger – A single station to replace Station 1 and Station 3
does provide good coverage as shown in Figure 29. There are, however, three problems with this
station merger:
The Sand Hill Road developments no longer have adequate fire department coverage
from the new merged location.
There is less fire station overlap in the high-demand downtown area. This can be
compensated for by having two engines at the new merged station, but then it makes
little sense to merge the stations in the first place.
It may not be possible to make good response times from the merged station to areas
along the northern border of Palo Alto.
Theoretically this merger looks good for coverage, but fails to achieve good results when
other factors such as performance and reliability are considered.
Figure 29: Coverage - Station 1 and Station 3 Merger
Station 2 and Station 5 Merger – The potential to merge Stations 2 and 5 has significant
merit. A new merged station at the intersection of Arastradero Road and Hillview Avenue can
provide a similar level of coverage as the two individual stations do at their present locations. In
addition, the Foothill Expressway provides excellent access to the areas currently covered by
these stations. It is also an route to access Stanford University from Campus Drive. Likewise,
this location would provide excellent access to the research park (where Station 2 is currently
located) and the residential areas of Station 5. Staffing of this station, which could be a combined
station with a crew cross-staffing an engine and rescue is discussed later in this report.
Figure 30: Station 2 and Station 5 Merger – Possible New Location
Recommendation 20: Merge Station 2 and Station 5 at a new location near the intersection of
Arastradero Road and Hillview Avenue.
In this section we discuss the deployment of fire department resources to the various
stations throughout the city, which includes the services provided under contract to Stanford
University and the Department of Energy (DOE) at the Stanford Linear Accelerator (SLAC). The
results of this analysis and the deployment changes recommended by it were derived from the
risk and demand analyses presented earlier, and from our professional observations on changes
to improve the deployment of emergency response and the PAFD as an organization.
Overview
The PAFD is very traditional in its delivery system, which is to devote a majority of its
effort and resources toward fire suppression even when EMS service is mostly what it does. As
the incident data and analysis in previous sections show, EMS calls far outnumber fires and
structure fires occur with very low frequency. As this trend will continue into the foreseeable
future, a key for the PAFD (and the city) is to take the initial steps to adjust the system such that
EMS resources are increased. A question posed to us for this study was: ― can the fire department
use dynamic deployment and shift resources based on demand during different periods of the
day, week, or even month?‖
To some degree the fire department is already using a dynamic deployment strategy,
though not on a large scale. There is currently a 12 hour paramedic unit and weekday BLS
service is staffed by part time by hourly employees. As discussion in earlier sections revealed,
the fire department has a major problem with data collection and analysis, so opportunities to
dynamically deploy resources are limited because the department does not have the capacity to
analyze data to the extent necessary for system-wide dynamic deployment. Likewise, the
dispatch center is not yet capable of implementing such a system, in part because it has not yet
implemented the Medical Priority Dispatch (MPD) software already available.8 Still, there are
opportunities to make the system more efficient and use the potential savings to increase medical
service. As data collection and analysis improves, the fire department can expand its efforts and
make other changes.
The PAFD has not changed its approach to emergency service protection in many years.
The current MOU is overly restrictive in its requirements for staffing and few changes can be
made until language in the contract is changed. To this extent the city may be limited in the
changes it can make even when they make sense (operationally or financially). Another
limitation for the city is the size of the current fire stations, which are small (and in some cases
outdated) to accommodate more apparatus or personnel. Savings are possible if two stations are
merged, but the city would need to construct a new station to achieve the savings.
To fully understand the discussions in this chapter more fully, the following definitions
are provided:
Engine – An engine is used primarily to supply water at the scene of a fire. These units
are staffed by three personnel, including an officer. In Palo Alto, one of the three persons on an
engine is usually a paramedic, thus engines improve the medical response system by providing
advanced life support (ALS) medical care even before a transport unit arrives.
Type III Engine – A type III engine is a smaller version of the engine listed above. Its
pumping capacity is also less than an engine and its use is primarily for brush and wildland fires
and the unit is specially designed for off-road use. When staffed, the Type III engine has three
personnel, including an officer. The unit in Palo Alto is only staffed during the months of July
through October and only then for 12-hour periods.
Rescue – A rescue unit is primarily responsible for extrication at accidents and the
personnel augment the on-scene staffing at structure fires. PAFD rescue personnel are also
trained and equipped to provide technical rescue and hazmat services. This unit is staffed by
three personnel, including an officer. One of the individuals may also be a paramedic.
Ladder – The ladder is primarily responsible for performing search, rescue, and
ventilation services at a structure fire. It is also capable of performing vehicle extrication and
supporting the rescue at the scene of a technical rescue incident. The fire department has only
8
For unknown reasons the MPD system has not be used.
one ladder (Station 6). This unit is staffed by three personnel, including an officer. One of the
individuals may also be a paramedic.
Medic or ALS (Advanced Life Support) Unit – These terms are used interchangeably
to describe a two-person medical transport unit staffed by two paramedics. The PAFD staffs two
of these units now, one 24/7 the other for 12 hours/day.
BLS (Basic Life Support) Unit – BLS describes a two-person medical transport unit
used to provide intra-facility transports and handle non-life threatening medical calls when an
ALS unit is unnecessary. This unit is staffed by two, non-safety, emergency medical technicians
hired under contract. The fire department maintains a list of these individuals and schedules them
as needed.
The demand for services in Station 8‘s area is too low to justify the cost and other
options may be available
The low call volume and workload for Rescue 2 suggest that it be cross-staffed by the
engine crew with the rescue staffing eliminated
Fire stations 2 and 5 can be merged as they have significant overlap not justified by
current or projected demand
Minimum shift staffing can be reduced slightly and still allow the second EMS rescue
to be staffed 24/7 instead of the present 12 hours
To maintain their usefulness and longevity, Stations 3 and 4 need capital upgrades
A protocol-based dispatch system methodology is needed for EMS calls such that fire
units are only dispatched to high priority medical calls, not every routine medical
situation
EMS can be more efficient if ALS rescues were staffed by one paramedic and one
emergency medical technician (EMT) instead of two paramedics9
9
County policy allows medic units to be staffed with one paramedic and one EMT.
The current fire-based ALS transport and BLS system is recouping a significant
portion of the total cost and the ALS system now in place should be maintained, at
least for the near term
The level of service for ALS service in Palo Alto (based on response times) is better
than going with the County transport provider
The fire department‘s overtime budget can be significantly reduced by eliminating
fulltime staffing on the rescue
As part of this study we reviewed the draft report of the County‘s emergency services
conducted by the Local Agency Formation Commission of Santa Clara County (LAFCO).
Studies by LAFCO are conducted every few years and they are important for Palo Alto because
their focus is to improve regional efficiency. Findings reported by the study with significance to
Palo Alto were:
Bays/ Year
Station Location Constructed First-Line Apparatus Minimum Staffing
1 301 Alma Street 3/ 1965 1 Engine 3- 24/7
1 Medic Unit 2 - 12 hours/ day
1 BLS Unit 2 (contract civilians)
weekdays
2 2675 Hanover Street 3/ 1965 1 Engine 3 – 24/7
1Rescue 3 – 24/7
1 Medic Unit 2 - 24/7
3 799 Embarcadero 2/ 1942 1 Engine 3 – 24/7
4 3600 Middlefield Road 2/ 1954 1 Engine 3 -24/7
5 600 Arastradero Road 2/ 1962 1 Engine 3 – 24/7
10
Staffed 12 hours per day on overtime
In addition to the above apparatus that are staffed either full or part time, PAFD also has
in its inventory two reserve engines, one reserve ambulance, and several technical rescue trailers
and technical support units, including a breathing air support vehicle. These units are necessary
to support the fire department‘s operation and mission and are considered essential assets.
Further analysis of the current delivery model is possible by looking at the level of effort
provided by the various resources deployed. Table 24 shows the minimum staffing and level of
effort for the various units.
Table 24: Level of Effort, PAFD
The purpose of the above exercise is similar to that of a company evaluating whether its
efforts are directed to the product offered and those which are being sold the most. If, for
example, refrigerators are the area of highest sales, it is doubtful that a company would build (or
have more sales staff) in the television department.
From the table above one can see that a majority (75 percent) of the Department‘s effort
is directed to fire suppression, even though medical calls occur with much more frequency than
fires, especially serious structure fires. Later in this section we use this information to
approximate the time spent by the fire chief and deputy chief managing the EMS system by
equating their workload to the EMS level of effort. Our rationale is because these individuals
would likely spend their working time in relative proportion to the resources and personnel of the
various units; e.g., more personnel typically results in more personnel issues.
The current MOU stipulates the number of units and personnel that must be maintained.
Currently the minimum staffing is 29. The minimum staff number does not include positions to
staff the 12-hour ALS unit, the shift battalion, or the BLS unit. Though we understand the
concern of the union to maintain an adequate staff to maintain safety, we disagree that the total
minimum staff should be in the contract. For the purpose of ensuring employee safety,
establishing the minimum number of personnel on an engine or ladder might be reasonable since
these are related to the tasks performed, and ultimately to the safety of the personnel working.
However, the city should never agree to a minimum staffing requirement that establishes the
total force as this equates to establishing the level of service provided.
Emergency Response
PAFD resources are assigned to eight response districts and units are dispatched outside
of their assigned district when the incident requires multiple units, or when the unit assigned to a
particular district is unavailable. As depicted earlier, seven of the districts have fire units
available 24/7. Station 8 is only staffed during the months of July through October.
Fire suppression units are all staffed with three persons to include one captain, one
equipment operator and one firefighter. Because ALS first response by fire units is provided, one
of the three individuals is usually a paramedic. However, the current policy is to staff each ALS
unit with two paramedics, thus there is not always enough paramedics such that every fire unit
also has a paramedic. With the exception of Engine 8, which is a Type III engine designed for
wildland fires, all of the other engines are designed for structural firefighting (Type I engines).
The fire-based system with paramedics assigned to suppression units is a common approach
throughout the region.
For medical calls and transport, two ALS units are available. The city‘s primary medic,
located at Station 2, is staffed 24/7. A second ALS unit located at Station 1 is also staffed by two
paramedics, but only for 12 hours each day. The PAFD pays overtime to staff the ALS unit at
Station 1. A third medic unit (reserve) is also available when it is not being used as a
replacement vehicle for medic 1 or 2. When available, the medic at Station 1 is staffed by the
engine crew during periods of high demand. The fire department maintains the right to provide
medical transport service separate from the County by virtue of maintaining its Legislative 201
rights.
Operational Changes
Based on the station location and response time analysis and the workload analysis of
current units, changes to the city‘s deployment model are feasible. There are limitations because
some of the changes can only be made if a new station is constructed such that Stations 2 and 5
can be merged.
Rescue 2 – Rescue 2 is a fully staffed unit whose primary responsibility is to perform
extrications at vehicle accidents and other situations where individuals become trapped. It is also
an additional resource on structure fires. At structure fires rescue personnel perform ventilation,
search and rescue, and other support activities that allow the fire to be extinguished effectively.
The rescue in Palo Alto is also the city‘s hazmat team and recently the unit was designated to be
part of the state‘s hazmat response system.
The functions performed by the rescue are important. The problem is the number of calls
for rescue and hazmat services are extremely low. In particular, structure fires are infrequent as
shown in the risk and demand section of this report. The bottom line is there is just not enough
activity to justify an independently staffed rescue in Palo Alto. If operated on a regional basis
where the cost is shared by several communities, the call volume might justify its cost as a
fulltime resource.
A better approach is to cross-staff the rescue with the engine co-located in the same
station. To do this the on-duty crew would handle calls as they occur in what is generally called
‗handle first emergency‘. For example if a medical call occurs, the engine would respond. If the
call is a structure fire in another district, the rescue would respond. Response policies would
need to be revised. To make the cross-staffing approach work effectively, the cross-staffed
engine/ rescue could be staffed by four personnel, not three, to include one officer, one
paramedic/ firefighter, and two firefighters.
Recommendation 22: Eliminate the fulltime staffing of Rescue 2 and cross-staff it.
If the engine and rescue at Station 2 were cross-staffed with three personnel, the
equivalent of 12 FTEs could be eliminated with an estimated savings of $1.38M. The savings
would result from a decrease in overtime. If the unit were staffed with four personnel instead of
three, the savings would obviously be less.
There is also an opportunity to work with other departments in the county to examine the
feasibility of regionalizing the services provided by the rescue. Hazmat and other functions
performed by the rescue are needed elsewhere as well and the city could further save if the unit‘s
cost were shared.
Recommendation 23: Begin discussions with surrounding agencies about sharing rescue and
hazardous materials response capabilities. The outcome of discussions would determine the
direction of rescue services when Stations 2 and 5 are merged. It would also determine the
footprint of the new station depending on the location of the rescue. Mountain View and Menlo
Park are two principle agencies to include in the discussion.
Staffing a Second ALS Unit – Demand for EMS far exceeds that for fire suppression,
yet the resources available to handle medical calls is much less. The analysis also shows that
medical calls will increase over the next ten years at a rate higher than for other calls. To
improve EMS capabilities the 12 hour ALS unit should be staffed 24/7 such that the PAFD is
able to handle almost the entire call volume for medical calls, especially during weekday hours
when the BLS unit is also staffed. Reportedly, an unfinished EMS study by another consultant
was to make a similar recommendation. The unit should continue to be located at Station 1
because it is an area of very high demand.
The PAFD presently cross staffs a third ALS unit with the engine crew at Station 1 when
the reserve ambulance is available for use. After the second ALS unit is staffed fulltime, a 3rd
unit can continue to be cross staffed. Because fulltime ALS units would be located at Station 1
and 2, the cross-staffed unit should be located at another station. Station 7 would be a good
choice for the reason that it has a low call volume and a paramedic is needed at Station 7. To
place the unit at Station 7 however, requires that SLAC agree that the 20-minute rule for
backfilling Station 7 be modified.
Beginning January 3, 2011, the 24 hour medic van was moved to station 1 and the 12
hour medic van was moved to Station 2. Engine 2 now cross-staffs Medic 3 afternoon. This was
done to alleviate the demand on Engine 1 in District 1.
PAFD expressed concerns about the Station 7 proposal because without the modification,
another engine would have to backfill for any response. Also, an ambulance will not fit in the
station; only the engine fits which may require facility modifications. Staff felt Station 3, 4, or 5
would be better locations for the cross staffed ambulance but this would ignore the low
utilization of Station 7 personnel and equipment.
Recommendation 24: With the savings of cross-staffing engine and Rescue 2, staff a second
medic unit 24/7 and continue ALS transport. Also, staff a third medic (weekdays and during
expected high call periods) using overtime and cross staff the third medic during non-peak times.
The anticipated incremental cost to maintain 24/7 staffing is $427,400.To calculate this we
determined the cost to staff the unit 24/7, then subtracted the overtime to staff the unit for 12
hours/ day.
Merging Stations 2 and 5 – Engines 2 and 5 both have low call volumes and work-
loads. As the response time analysis shows, if the stations were to be merged response times to
the area would still be adequate. The general area of Miranda Avenue (Foothill Expressway)
near Hillview Avenue is a good location to consider because it has excellent access across
Junipero Sierra Boulevard toward Campus Drive to access Stanford University. It is also
excellent to access Arastradero Road by way of the Parkway toward to the area covered now by
Station 5.
By merging these stations one engine company could be eliminated. The rescue, if it
remains solely a city resource, should be located here along with one engine. Depending on the
move made concerning Station 8, the Type III engine could also be located at this facility. Based
on data we reviewed, it is estimated that the call load would be approximately 1400 emergency
responses with an emergency workload of approximately 700 hours – very manageable. This
would be similar to Engine 1 now, which responds to 1614 emergencies and has an emergency
workload of 698 hours.
Recommendation 25: Merge Fire Stations 2 and 5. A suitable location to maintain excellent
coverage and improve efficiency is in the vicinity of Miranda and Hillview Avenues. Because
planning for a new station would take several years, the city would be able to make small
changes such as cross-staffing the engine/ rescue and staffing the second medic while it evaluates
the larger question of station consolidation.
Approximately 11 FTE positions could be eliminated by merging Stations 2 and 5. This
would include three captains and nine firefighters, or four personnel on each of the three shifts.
Based on the average salary and benefits for captains and firefighters the savings are estimated at
$1.35M.
Recommendation 26: When Stations 2 and 5 are merged, eliminate one engine and continue
to cross-staff one engine and the rescue. As an alternative, the rescue could be independently
staffed if the unit is funded by the multiple jurisdictions such that it becomes a regional resource.
Station 8 – A question posed by the city for this study is whether staffing Station 8 in the
west Hills district area makes sense, and what other options might exist?
Fire Station 8 is located in a rural area west of the city. It is staffed with three personnel
using a Type III engine that is kept at Station 5 when not being used. Station 8 is only staffed
during the ‗assumed‘ high-fire season and then for only 12 hours each day from July to
November. In all of 2009, Engine 8 responded to only 17 calls and the total workload for these
calls was only 12 hours. The cost in overtime to staff Station 8 is $188,000.
The cost to staff Station 8 was previously shared by the city of Los Altos, which paid
Palo Alto $50K per year. A new station was constructed by Los Altos nearby which led to Los
Altos deciding not to continue paying the cost. Although we did not analyze the location of the
Los Altos station relative to the situation in Palo Alto, this is probably a ‗lost opportunity‘ to
improve regional efficiency. Considering the cost versus the return, questioning whether Station
8 should be staffed is justified.
The problem with staffing this station is not just cost, but also logic. Assuming that a fire
will occur only during the months July through the end of October makes no sense. Serious
brush fires can occur any time of the year. Also, there are only about 10 high-fire days each year
and the days are known by state officials because they evaluate the conditions. Another problem
is that medical emergencies occur with much greater frequency than fires. If the unit is only
staffed 12 hours per day over three months, is it likely the unit would be available to handle a
cardiac arrest? Not if the unit is available only 12 percent of the time (1080/8760).
A better and more cost efficient way might be to install infrared technology that would
allow the entire wildland region to be monitored for a potential fire 24/7. Such technology is
available and has been used for some time. The initial cost is unknown, but the city could
probably partner with Stanford University‘s engineering department to develop the system. It
would be an excellent way to improve efficiency and the project would likely be embraced by
the university as a good project for engineering students. The city might also be able to involve
other agencies such as the DOE, the primary tenant at SLAC.
Recommendation 27: Partner with Stanford University to develop and install infrared (IR)
technology in the west Hills area in lieu of staffing Station 8 only part of the year. Include
residents of the west Hills area in the planning process.
We also considered whether other agencies have interest in sharing Station 8 such that the
cost could be reduced to the city and service level hours extended. For this we contacted Cal Fire
officials to determine possible interest because they have a nearby station with two engines. Cal
Fire would consider moving one of the two engines already located at the station near Skyline
Boulevard south of Los Trancos Road. Options and the estimates to provide an engine at Station
8 for five months (24/7) were:
Depending on the direction taken for cross-staffing a rescue and engine at the new station, the
same crew could also cross-staff the Type III engine because the wildland fire workload would
be minimal.
A third possibility is to involve the current members of the CERT team and train them as
an auxiliary force to staff Station 8. The city has many citizens participating in the CERT
program and they are very active. This would require a contractual modification and staff
expressed some concern because of the average age of CERT participants.
The most appealing and potentially cost-saving initiative may be to share the cost of a
fire station with Menlo Park, which has a station less than one mile from Palo Alto Station1.
Certainly there are obstacles to this because Menlo Park is in San Mateo County, but it should be
explored. There may be other opportunities as well with Los Altos and Mountain View.
Recommendation 30: Hold discussions with Menlo Park to evaluate the possibility of sharing
the cost of operating Station 1 in Palo Alto with Station 7 in Menlo Park. Discussions should
also be conducted with Los Altos and Mountain View.
are 24/7 operations, the staffing factor (multiplier) determines the level at which to staff the
department based on its daily staffing requirements. For overtime the city wanted to know the
‗break-even point‘ where it makes financial sense to hire additional firefighters rather than to pay
current staff overtime.
Staffing Factor – To determine the PAFD staffing factor we examined leave data for all
fire personnel for an entire year. The staffing factor is based on the total number of hours in a
year (8,760) divided by the average number of available hours for a single firefighter. Palo Alto
firefighters are paid for 2,912 hours per year, but because of holidays, vacation, sick leave, other
time off, and other duties such as training, a firefighter is not really available to work all 2,912
hours.
On average, each firefighter was absent an average of 405 hours during FY2009–2010.
Thus, a firefighter is only available to work 2,507 hours. The staffing factor to cover one on-duty
position 24x7 (8,760 hours per year) is 3.49; that is, it takes 3.49 firefighters (8,760 ÷
2,507=3.49) for every position filled 24/7. Because the absence data is provided by position
classification, we were able to determine the staffing factor for the positions of battalion chief,
captain, and firefighter.
The number of FTE positions necessary to staff the PAFD does not include those filled
through overtime such as Medic 1 and Station 8. Table 25 shows the staffing factor based on the
data we analyzed.
Table 25: PAFD Staffing Factor Analyses, FY2009–2010
A minimum of 105 FTE position are needed to staff the current shift fire system
comprising one battalion chief, seven engines, one ladder, and one rescue 24/7. The overtime
budget appears much higher than it should be given that 105 FTE positions is the authorized
strength for the PAFD. We suspect that the actual staffing factor is probably higher than we
analyzed, possibly because not all absences are being recorded. If station personnel are assigned
temporary or light duty, or to outside training and these are not included in daily absence records,
the fire department could not reasonably know how many firefighters (or overtime) are needed
for its situation.
To determine the staffing necessary for the changes we recommend in this study, leave
data provided by the Department and our calculation of the staffing factor were used. Leave data
used in the analysis is provided in Appendix B.
Overtime Analysis – Some overtime will always be incurred because of daily staffing
fluctuations that require more firefighters to be off above the norm. Overtime often costs less
than carrying more firefighters—the cost of overtime may be less per hour worked if benefits are
considered. The tipping point depends on whether the overtime is earned by experienced higher
paid firefighters vs. full time entry level firefighters. The results of the analysis below will
probably be surprising to those who hold the widespread belief that overtime significantly drives
up costs. In fact if you don‘t spend enough on overtime, and have extra full time firefighters
above the number needed each day, it cost more, not less.
According to the Department, 36 positions per platoon are available when each shift is
fully staffed. However, because of temporary assignments, disability, and vacancies it is unusual
that all 36 positions are actually filled. In addition, the labor agreement provides that a maximum
of four positions can be scheduled for vacation each day. As a result there are days when only 32
persons are on duty even if the shift is at full staff. This leaves only two positions above the
minimum needed to fill the 30 on-duty minimum slots. If there are other absences, or if the shift
is understaffed, overtime will be required.
To determine how often overtime is used to maintain the minimum on-duty staffing, an
analysis of overtime use was performed for FY 09-10. The Department provided a data file that
specifically identified the overtime used for the on-duty minimum staffing. The analysis
involved identifying the overtime hours needed each day to fill the on-duty minimum
requirement. Overtime for Medic 1 and Station 8 was not included in the analysis because of the
limited and part time nature of these services.
Table 26 shows the number of overtime shifts and the number of days that those shifts
occurred.
Table 26: Distribution of Overtime Shifts, July 2009 to June 2010
Number of 24 Hour
Overtime Shifts Percent of Total
per Day Number of Days Days
0 13 4%
1 30 8%
2 41 11%
3 63 17%
4 68 19%
5 58 16%
6 43 12%
7 30 8%
>7 19 5%
Total 365 100%
In 2009, $4.0M was required for overtime (Appendix C). A key financial question is
whether there is a cost benefit to hire additional personnel staff instead of paying overtime. To
determine this various scenarios were tested. Regardless of the scenario tested, hiring additional
firefighters did not always save more overtime than what it cost to pay the firefighters.
The reason that hiring more firefighters is not cost beneficial is that the total salary and
benefits cost is slightly higher per hour than the overtime costs. The number of hours worked and
the 54% benefit costs significantly raise the cost of a new hire compared to using overtime,
which adds only 50% to the hourly salary cost and only an incremental 1.8% in benefits. Based
on the salary and benefits costs of a new firefighter and the number of hours that would be
available compared to the cost of overtime to fill those same available hours, it is generally more
cost beneficial to use overtime.
If new firefighters are being hired all the time, such a hiring strategy might produce some
savings. However, as firefighters progress to higher pay grades each year the cost benefit of the
overtime advantage decreases as the firefighters‘ salary increases. As years go by, any initial
savings differential begins to disappear.
Figure 31, shows the total costs of one firefighter available per day at different pay levels
compared to the cost of the overtime at step 5 based on the number of overtime hours. When new
firefighters are hired (Assumes that the staffing factor is at 3.5), there is a point (8,368 hours)
when hiring new firefighters is slightly less expensive than paying overtime. However, as this
group of firefighters remains with the Department, the cost of this group at salary steps 3 and 5 is
always higher than the cost of overtime, and as a result there is no breakeven point.
$550,000
FF EMT Step 5
$500,000
FF EMT Step 3
$450,000
Annual Cost
FF EMT Step 1
Overtime
$300,000
6,000
6,100
7,500
7,600
7,700
7,800
7,900
8,000
8,100
6,200
6,300
6,400
6,500
6,600
6,700
6,800
6,900
7,000
7,100
7,200
7,300
7,400
8,200
8,300
8,400
8,500
8,600
8,700
8,760
Annual Hours
Although the above analysis suggests that overtime saves money compared to hiring
additional firefighters, the analysis does not account for the impact on morale and fatigue if the
existing firefighters must continually work overtime because of staff shortages. Thus, the fire
department must also balance the cost of overtime with the reality of overtime on firefighters and
its operation. Also, for a major emergency requiring a surge of personnel on duty, the bodies will
not be there if overtime has been totally depended on for staffing.
EMS Delivery Model and Cost Analysis – The fire department uses a three-tiered
system to handle medical calls. The primary response is by engine companies who are the first
responders. The second tier is the ALS transport service, which is provided by two paramedics.
The third, and essential tier, is the BLS transport unit that is staffed by civilian Emergency
Medical Technicians (EMTs) during weekdays.
EMS generates approximately $2.1M annually and response times by city medic units is
much better than the 12-minue goal for American Medical response (AMR) required by the
County. A change to AMR as the transport provider would likely raise concerns among the
public as an unacceptable service-level reduction. It should be noted that Santa Clara County has
entered into a new vendor contract for ALS transport services with Rural Metro beginning July
1, 2011.
Although the model used by the PAFD is generally a good one, there are problems. First,
there is not enough management oversight for the program and the current EMS coordinator is
overwhelmed trying to keep up with the workload, especially training and the Continuous
Quality Improvement (CQI) program. There are also issues with how calls are dispatched and
handled because current policy requires sending a fire unit and an ALS rescue on every medical
call, regardless of how minor. Additional ALS transport capabilities are needed, but if the
dispatch protocols are not changed, adding another fulltime ALS unit might not be enough. And
though not a critical problem, it is our opinion that shift battalion chiefs should be required to
have (and maintain) ALS certifications, for the reason that the PAFD‘s main effort is toward
medical response, not fire suppression.
First Response: A primary reason for fire departments to begin delivering medical
response services is their emergency units were deployed in smaller response districts, therefore
arrived at emergencies more quickly than centrally located ambulance services not able to meet
recommended response standards. Most national medical standards seek a staged system with the
minimum standard being that BLS first-response arrival within four minutes and ALS within
eight minutes, 90 percent of the time. Many communities do not meet this goal but Palo Alto
does, in most cases.
California Health and Safety Code Section 1797.201 (referred to as 201 rights) allows
cities, that provided care prior to June 1, 1980, to continue and retain administration of pre-
hospital care services, however the County Health Officer maintains control of medical
standards. Palo Alto is one such City, providing both ALS and BLS transport. The fire
department generally has enough paramedics to staff its units, however, there are issues with
maintaining the necessary paramedics for all fire units to be ALS staffed all of the time.
Cities, including Palo Alto, question the necessity of sending a 20-30 ton fire engine or
ladder truck to medical emergency. Smaller units would be more efficient and less costly than
fire engines, but additional staffing would most likely be necessary or the smaller units would
have to be cross staffed. Some agencies are migrating to a cross-staffing model based on the fact
60 percent or more of their calls are medical in nature, therefore the risk of increased response
time due to concurrent fire related emergencies is small and the change is justified.
For Palo Alto, which has a real need to improve its data collection and analyses system
before it can become more sophisticated in its deployment strategies; the current fire-based ALS
first response system where first-line units respond with paramedics and medic units provide the
transport is a good approach. As the fire department becomes more effective with analyses, and
the organization begins to make other changes, then cross-staffing of light-duty response units
can be explored. The facilities issue of fire stations that are already too small must also be
addressed such that small response units could be added to stations, in addition to the fire units
there already.
Management of EMS: Oversight of the EMS program is provided by the deputy chief
(operations) with day-to-day coordination by a civilian coordinator. Clearly, the three-tiered
EMS system in Palo Alto is complex and there is a need to add additional oversight with the
current coordinator, who is not uniformed, providing program support. A major concern of the
current system is that the coordinator, as a civilian, is unable to respond to emergencies, or
participate as a member of the incident command team. Given that a major emergency in Palo
Alto could well be a mass-casualty incident, the lead person for EMS should be a battalion chief.
Recommendation 31: Maintain the current fire-based EMS first response system and add a
battalion chief to manage its operation. The organizational structure we recommended in
Chapter II combines the responsibilities for training and EMS under a single battalion chief.
There is also a problem with the current dispatch protocols that require a fire unit to be
dispatched on every medical call, even minor ones. The fire department‘s policy is also to send
an ALS unit to every medical call even when the BLS unit is available. These are unusual
policies that make little sense when efficiency is important, especially during busy periods. It
was also explained that emergency dispatchers do not use the Priority Medical Dispatching
(PMD) software that is already installed in the CAD program. Properly used, PMD with pre-
arrival instructions can and does save lives. It also aids in making certain that the right resources
are dispatched.
Recommendation 32: Develop a protocol based dispatch system methodology for EMS calls
such that fire units are only dispatched to high priority medical calls, not every routine
medical situation. Also, implement the medical priority dispatching system already available in
the CAD software. Making these changes will lessen demand on the total system, particularly
during peak periods, and it will improve system efficiency.
Revenue and Cost Analysis: A question for this study was whether the cost to deliver
EMS transport by the fire department is fully recovered through the fees charged to patients. The
PAFD does not specifically budget for its ALS ambulance service but does have a separate
budget for its BLS ambulance service.
To estimate the salary and benefits costs for the two paramedics that staff Medic 2 (24/7),
we used the firefighter staffing factor and the salary and benefit costs for a paramedic at
$164,463. For two on-duty positions, at least seven paramedic staff are needed to provide the
staffing, and the salary and benefits cost for the seven paramedics is $1,151,241. In addition we
added the overtime costs already budgeted in the Advanced Emergency Medical-Paramedic cost
center. The FY 09-10 costs from the rest of the Advanced Emergency Medical-Paramedic cost
center were used, which also included vehicle replacement costs of $31,070 and vehicle
maintenance costs of $66,057.
The current Advanced Emergency Medical-Paramedic cost center has salary and benefits
for portions of various positions, such as a Battalion Chief, EMS Coordinator, Firefighter
Paramedics, Fire Apparatus Operators, Firefighters, Operator Paramedics, a Deputy Fire Chief,
and Administrative Associate IIs. Because the specific costs were not identified, none of these
other costs were included. However, a position like the EMS Coordinator would increase costs if
these costs were fully accounted within the ALS budget.
Table 27 shows the revenues for both ambulance programs and the combined estimated
costs just to provide the services. The costs do not include portions of salaries and benefits from
other positions that provide indirect support to the programs or costs associated with
departmental overhead and other costs. The costs just to provide the ambulance services are
slightly less than the revenue by $42,650.
Table 27: Estimated Cost of Service and Cost Recovery, Palo Alto EMS
Revenues
ALS Fees $1,986,026
BLS Fees $234,489
Total Revenues $2,220,515
Expenditures
BLS Ambulance Program $91,864
Medic 1 Overtime $427,392
Medic 2 Salaries & Benefits 1,231,807
Other advanced emergency medical-paramedic costs $426,802
Other staffing costs Unknown
Total Expenditures $2,177,865
Difference $42,650
If the cost to manage the EMS program, including oversight by the EMS Coordinator, is
included, the costs become greater than the revenues. The EMS Coordinator spends 100 percent
of her time coordinating EMS programs. We also added portions of the fire chief and deputy
chief salaries from the apportionment of the levels of service earlier in this section. The
additional cost can be calculated as follows:
Full cost of EMS Coordinator/ benefits = $161,652
Partial cost (25 percent) of Fire Chief/ benefits = $67,075
Partial cost (25 percent) of Deputy Fire Chief/ benefits = $59,179
Total additional cost = $287,906
The cost for just these salaries (and benefits) increases the total cost, and as a result, the
costs are greater than the fee revenues by $245,256. Even if the unrecovered cost was to be
slightly higher than $245,000, it is likely that citizens because of other factors would opt for the
ALS transport provided by the PAFD over the longer response times achieved by AMR.
Additional revenue can be generated if a second ALS transport unit is staffed 24/7 instead
of the current 12 hours.11 However, additional staffing costs would also be incurred, and the
increase in revenue might not offset the additional costs. To calculate the total cost for EMS
transport more precisely, the fire department should also include such things as utility costs for
the space used at the various stations, administrative overhead, and other indirect expenses. An
analysis of this level is beyond the scope of this project.
Our review also found a critical need to improve the accounting system for ALS and use
a similar enterprise accounting system for ALS that is now used for BLS. Reportedly,
expenditures and revenues can be tracked using the Advanced Emergency Medical-Paramedic
cost center. In this way all costs, including portions of the salary and benefits for management
staff as we estimated earlier can be factored in.
EMS Fees and Collection Rates: The City has a variety of fees that it charges for
transporting patients via the City‘s ALS and BLS ambulances. The fees range from a single BLS
transport at $450 to an ALS II transport at $1,595. Other EMS charges involve mileage, special
event standby, scheduled BLS transports, night transports, and oxygen. Appendix D shows the
City‘s transport fee schedule.
Although the City has a set fee schedule, the collection rate for the fees charged averaged
only about 53% in FY 09-10. Based on information provided by the City‘s billing and collection
11
It is estimated that a minimum of 500 transports per year would be achieved if the second ALS were staffed
fulltime. Assuming the current mix of ALS and BLS transports, the estimated revenue from these additional
transports is $352,500.
vendor, the City charges totaled $4,042,274, but the collections only totaled $2,146,333. The
average payment was $705. The amount collected primarily depends on the type of insurance.
For example Medicaid and Medicare patients had collection rates of 10% and 35%, respectively,
while patients with auto insurance, Kaiser, or private insurance had collection rates of 92%, 67%,
and 78%, respectively. The City‘s ability to generate revenue from EMS transports is based on
the mix of patients and their insurance coverage. In FY 09-10 45% of the patients were on
Medicaid and Medicare, while another 47% paid with auto insurance, Kaiser, or private
insurance. Often the transport fee covered by insurers is limited or has a maximum amount, and
such limitations can affect the City‘s ability to collect additional revenue when it raises its fees.
What is not covered by insurance must then be paid by the patient or written off depending on
the City‘s collection policies regarding these types of fees.
As previously discussed, when fee revenues do not cover the costs of EMS and the
related transport services, the City provides the funding to help offset the costs of the ALS and
BLS ambulance program. In the future, insurance, demographic, and personal economic changes
might play an important role in the City‘s ability to maintain and generate additional transport
revenues.
EMS Transport Options: If, when a thorough cost accounting is completed, revenue
does not justify transport by the fire department, the city can consider other options. It could
contract for transport to a private firm under a franchise format. Or, it may be able to establish a
separate contract with AMR under the County agreement with higher service-level standards to
include dedicated transport units for the city to provide faster response times than required by the
County. These and other options and a full review of the entire EMS operation were to be
addressed in a study by another consultant; however, the study was never completed. To explore
these opportunities and fully review the operation of EMS such a study should be completed.
Recommendation 33: For the present time, maintain the EMS transport system provided by
the fire department. An estimated 85 percent of the cost for EMS transport is being recovered by
ALS and BLS fees.
Recommendation 34: Conduct a thorough analysis of the EMS operation, including transport
options. There will be limitations to what the city can do because EMS policies are directed
under the authority of the County Health Officer, and thus a full review is needed.
Recommendation 35: Following a full EMS review, decide whether to replace the current
fleet of ambulances, or consider giving the ambulance fleet to the vendor if services are
contracted. If the city continues to provide transport, replace the ambulances beginning with
the unit having the highest maintenance costs.
Another need is to add a paramedic to the engine already located at SLAC such that one
of the three personnel on the engine is always a paramedic. Personnel at SLAC have concerns
that a paramedic is not available on the engine when the most serious hazard at SLAC is a
potential electrocution due to the high-voltage used. Even minor electrical shocks cause
arrhythmias that quickly cause cardiac arrest. If addressed quickly, the arrhythmia can be treated
by a paramedic.
Recommendation 36: Assign a paramedic to Station 7 such that Engine 7 can be operated as
a paramedic first-response engine.
Battalion Firefighter/
Station Units Chief Captain Lieutenant Paramedic Firefighter Total
1 1 Engine 1 2 3 6 12
1 ALS 3 3
2 1 Engine 1 2 3 6 12
1 Rescue
1 ALS 3 3
3 1 Engine 1 2 3 6 12
1 ALS (Peak)
4 1 Engine 1 2 3 6 12
5 1 Engine 1 2 3 6 12
1 Type III
6 1 Engine 1 2 3 6 12
1 Ladder 1 2 9 12
1 BC 3
7 1 Engine 1 2 3 6 12
8 Eliminated
Totals 1 BC 3 8 16 27 51 105
7 Engines On-duty
1 Ladder 24/7= 26
1 Rescue On-duty
(peak) = 28
2 ALS Units
(24/7)
1 ALS Unit
(Peak)
The anticipated savings if the immediate changes are implemented are estimated at
$1.57M. These include cross-staffing the rescue and engine at Station 2 and eliminating the
overtime for Station 8. The savings would be somewhat offset by the additional expense of
$425,000 necessary to staff a second medic unit 24/7 instead of 12 hours per day.
There would also be some additional overtime to staff a third medic unit during peak-
demand hours. We estimate this to be about eight hours each weekday, though some of time may
be during major athletic events at the University. There are also potential savings if, as we
recommend, two of the three officers assigned to each station are lieutenants instead of captains.
Each position changed from captain to lieutenant would save approximately 10 percent of the
salary and benefits of a current captain.
Table 29: Deployment Changes – Long-Term Opportunities (Stations 2 and 5 Merged)
Battalion Firefighter/
Station Units Chief Captain Lieutenant Paramedic Firefighter Total
1 1 Engine 1 2 3 6 12
1 Medic Unit 3 3
2 1 Engine 1 2 3 6 12
1 Rescue
1 Medic Unit 3 3
1 Type III
3 1 Engine 1 2 3 6 12
1 ALS (Peak)
4 1 Engine 1 2 3 6 12
5 Eliminated
6 1 Engine 1 2 3 6 12
1 Ladder 1 2 9 12
1 BC 3
7 1 Engine 1 2 3 6 12
8 Eliminated
Total 1 BC 3 7 14 24 45 93
6 Engines
1 Ladder On-duty
1 Rescue 24/7 = 23
2 ALS Units On-duty
(24/7) (peak) =
25
1 ALS Unit
(Peak)
Further savings would result if Stations 2 and 5 are merged. The anticipated savings of
reducing the minimum staffing to 23 personnel on-duty instead of 26 under the first model is
approximately $1.35M.
Fire Prevention
Given the city is built out and population growth will be minimal, the fire department
does not need to add resources in the next few years, especially if prevention activities are
maintained at a high level.
To maintain excellent fire safety and to keep from having to add firefighters as the city‘s
building stock gets older, the city should continue its efforts to increase the number of buildings
that are sprinklered. There is also a need for firefighters to be more active in prevention
activities, not just available after a fire starts. The work of this division is critical to the mission
of the fire department.
Staffing – Fire prevention is currently managed by an acting fire marshal (fire captain).
Primary responsibilities include plan checks, new construction inspection, consulting with
developers and builders, building permit inspections, and inspections mandated by the State of
California Fire Marshal. It is also responsible for coordinating the inspections conducted by fire
engine companies. Assistance with plan checking is obtained using a former assistant fire
marshal at an hourly rate.
Prevention staffing also includes three fire inspectors, two hazardous materials inspectors
and one administrative associate II. One fire inspector position was eliminated during a recent
budget reduction. This division is also responsible for improving training and data management
systems and improving the Geographic Information Systems (GIS). However, these activities
were put on hold because of limited staffing. Because it has limited staff, plan checks are
sometimes referred to contract employees.
This division is also responsible for specialized hazardous materials inspections in
approximately 400 occupancies and is responsible to implement the recently adopted wildland
urban interface management plan (Foothills Fire Management Plan) adopted October 2009 by
the City in which inspectors place emphasis on inspections, enforcement and public education in
wildland urban interface areas for fuel modification and evacuation planning.
Clearly, the prevention division has a very high workload and is very important to the
department‘s mission. For these reasons it is imperative that the city fill the fire marshal position
as soon as possible. To expand the potential list of qualified candidates, the city should consider
civilianizing the position, which in fact, the just retired fire chief was interested in doing.
Recommendation 37: Fill the vacant fire marshal position. The position does not need to be a
uniformed position. In fact, by hiring a civilian fire marshal, a broader applicant pool can be
achieved. Another consideration is to share the fire marshal position with adjacent communities
to reduce cost and improve efficiency.
Fire inspectors are currently sworn uniformed employees and are permitted to work
overtime at fire stations, though not on their regularly assigned work day during the week. In our
opinion allowing inspectors to work overtime is not a good practice for the reason that it detracts
from their primary responsibilities. Fire management counters that it is a good policy because the
inspectors bring with them their expertise when they work overtime at the station. Management
also believes it helps inspectors to keep their firefighting and EMS skills up-to-date. It is
questionable whether inspectors working a few shifts per year would keep their skills current
when this is already a problem for other firefighters who work shift work.
The department has considered civilian inspectors but has not acted to incorporate them
into the department. The city should consider doing this when openings develop. Also, if police
and fire are ultimately merged into a hybrid public safety department, it could consider merging
all of the city‘s code enforcement activities under one division. Alexandria, VA, which we also
studied, has had such a system for some time and it is working effectively. Another option to
consider is to merge all city code enforcement and inspection personnel under one department.
Inspection Fees: Fees are charged for some inspections but the fee schedule has not been
updated in several years. The FM believes there should be fees for all services. One possibility is
to implement a business model for the code enforcement division such that the entire cost of the
operation is paid for by fees. A best-practice organization doing just this is Saint Paul, MN, a
department that we studied several years ago. At the time of this study the city‘s fire marshal
believes that it is covering about 75 percent of its operational costs, though he is not sure and
data is not be analyzed.
Recommendation 40: Have the fire marshal conduct a complete review of the entire fire
prevention division, including its goals, workload, and inspection frequencies. Include a
review of the inspections conducted by engine companies with a goal to increase the
contribution of on-duty personnel toward prevention and code enforcement. At the same time
review the fee structure for various services with a future goal of covering the cost of all services
through collected fees.
As the fire department increases its productivity in prevention overall, one area of need is
to improve the inspection and mitigation of ‗defensible space‘ hazards in the west Hills area of
the city. Such enforcement can easily be accomplished by on-duty firefighters and it reduces the
already low fire risk for the area covered by Station 8.
Recommendation 41: Improvement enforcement efforts of ‘defensible space’ in the west Hills
area.
Public Education: The PAFD does too little in public safety education with a bike safety
program (generally a police department function), school visitations and other ‗show and tell‘
programs. Reportedly the bike safety program is conducted by firefighters while on overtime,
which is unusual. There is also a poster contest for K-3 and the fire department reported that it
did 329 public presentations of one sort or another in 2009. Overall, the fire department‘s formal
public education program with target-audience programs falls short.
For a city of its size, Palo Alto should have at least one fulltime public educator on its
prevention staff. By doing so, the city and its residents could save by further reducing fire loss.
Good public education programs also reduce unnecessary alarms, which in turn improves the
efficiency of the response system.
The individual responsible for public education should be a civilian, preferably with a
professional background in education. We encourage the city to look at Rome/ Floyd County,
GA as a community with a first-rate program. Rome/ Floyd County‘s program has been in
existence for over 20 years and is exceptional. In addition to providing education and fire safety
training, its public educators along with the county have tracked the retention rate of students
from K-12. All schools are required to present the programs, which are developed by the fire
department‘s public education specialist.
Recommendation 42: Add a public fire educator to the code enforcement division.
Other considerations to improve prevention and risk reduction efforts going forward
include:
Evaluate all high-risk structures not protected by automatic extinguishing systems and
prepare recommended ordinances to retrofit these buildings. Particular attention
should be placed on buildings four or more stories in height. Try to implement the
required retrofits over the next three to seven years.
Establish and monitor a data base of inspections and repairs by private contractors to
fire protection systems by structure by occupancy, time and date.
Conduct public education seminars about the potential problems with ionization
detectors and the differences between ionization and photo-electric smoke alarms.
Training
Training for PAFD personnel is a real problem area and it is one area that needs to be
addressed as quickly as possible. The key issue for training is that there is only one training
officer and his plate is full trying to address the entire training needs of the organization and be
the principal training planner simultaneously. A one person training staff can work in a small
organization where fire suppression is the only service provided. It is not effective though for
Palo Alto, which has a need to provide EMS continuing education and training for special
services such as technical rescue and hazmat. The training facility is also outdated and should be
replaced, preferably with a modern, regional facility. To its credit the fire department is doing the
best it can with the resources available. For example, fire personnel recently constructed have
constructed a roof prop to train personnel in proper ventilation techniques.
PAFD training is coordinated by a captain. Half of the captain‘s salary is paid by the
city‘s utility department for the reason that the fire department is responsible to provide
confined-space training to utility employees. In addition to this responsibility the training captain
is also responsible for EMS training, wildland fire training, as well as other tasks such as self-
contained breathing apparatus (SCBA) fit testing, which is required. Much of the training
coordinated by the captain is mandated by federal and state regulations such as CALOSHA.
A major concern of the union is the training. It believes that training is not of
sufficient quantity or quality.
Hands-on training is insufficient to maintain skills. Hands-on training called ‗Friday
fires‘ were previously conducted; however the training center is no longer able to
support such training because it is outdated. Weekday training is also a problem
because demand is highest during this period.
FFs may not be getting enough exposure to structural firefighting, in particular the
use of SCBA during hazardous interior operations, search and rescue, etc. Clearly,
they are not getting actual experience because there are few structure fires. To
compensate for experience, the PAFD must increase its commitment to live-fire
training, possibly by hiring instructors on overtime.
Sunnyvale is the closest department with a live fire-training simulator and Palo Alto
uses this facility infrequently. There are few buildings in Palo Alto available for a
live-fire exercise.
The PAFD does not use its 24-hour schedule to maximum efficiency, which could
improve training availability. For example training is typically not conducted during
evenings or on weekends. The Department must change its culture such that training
(and other activities are conducted during evening and weekends, not just weekdays.
The truck company at Station 6 is used as the ‗training company‘ and its crew
supports the training officer. This is an excellent concept and one that has been used
successfully in other departments (Cincinnati). Though excellent, this approach does
not always achieve the best results because the shifts change each day, thus there is a
loss in continuity.
The recruit training program that is coordinated with other departments in the county
is an excellent approach. New recruit complete one week of introduction at the PAFD
then join recruits from other fire departments for 12 weeks of job-related training.
They then return to PAFD for additional training specially targeted to PAFD
operations.
The training calendar currently used is insufficient as a planning tool. There is a need
to understand the level of effort needed for training. The training division does not
know the total hours of instruction necessary for the various training needs. Without
such an analysis it cannot prepare a training plan, or prepare an annual budget to
include part-time instructors on overtime.
Recommendations to improve the fire department‘s training effort are:
Recommendation 43: Add a battalion chief to plan and coordinate the training program and
maintain the current training captain as the day-to-day trainer. The battalion chief will also
provide management oversight for EMS with the EMS Coordinator reporting to the battalion
chief.
Recommendation 44: Have the battalion chief and training captain conduct a complete audit
of all training needs and prepare a multi-year plan for training. Use the plan to develop the
annual training plan which should include training priorities, assignments, and overtime for
adjunct instructors when needed.
Recommendation 45: Implement a formal program for career and professional development,
including specific requirements for promotion to all officer ranks. Include educational
requirements and attendance to National Fire Academy, such as the Executive Fire Officer
Program (EFOP) as part of the requirements.
Recommendation 46: Work with other fire departments (and the county) to regionalize the
city’s fire and EMS training program. Establish a planning group consisting of city and
regional officials to plan for the replacement of the city’s outdated fire training facility. The
city should not build its own facility, but partner with multiple agencies to provide an adequate
facility.
Recommendation 47: Improve productivity of the training program by conducting more
training during evenings and weekends. As part of this effort, establish specific training goals
for each shift battalion chief and captain.
9-1-1 Dispatch
A review of the city‘s 9-1-1 dispatch center was not part of this study. Because 9-1-1
dispatch is so important to the efficacy of the fire and EMS response system, we do include a
brief discussion of it.
Palo Alto has combined police-fire dispatching system managed by the police
department. Before being under the police department, 9-1-1 dispatching was a joint operation
with the city of Mountain View. In the 1980‘s the city removed itself from the partnership in part
because it wanted more local control. This is an unusual move inasmuch as most communities
were beginning to regionalize emergency dispatch systems even in the 1980s as CAD systems
became more prevalent. The future for the city should be a return to a regional dispatch center, if
only for departments in North County.
The combined fire and police dispatch center handles roughly 150,000 incoming calls
each year. The center has four positions: two police dispatchers; two fire dispatchers; one call-
taker. Fire units operate on UHF radio bands while police operate on VHF. The frequencies are
linked to provide interface between police and fire units at the scene.
Plans in the county are for a 700MHz system and the city is looking at the possibility of
going to a county-wide system, which it should. Discussions are also taking place with the city of
Los Altos and Mountain View on technical upgrades and virtual consolidation such that Palo
Alto and these communities could operate under a single system for fire dispatching.
There is an opportunity to improve service delivery and efficiency if Palo Alto and other
jurisdictions combined to not only share stations but also to implement a boundary-less response
model such that resources from any jurisdiction would respond regardless of the political
boundary. This approach would more than likely result in more service-sharing opportunities and
improve efficiency because cities could share resources more than they do now under the
existing automatic-aid agreements.
Recommendation 48: Establish initial efforts with Los Altos, Mountain View, and Sunnyvale
to implement a boundary-less response network.
INTRODUCTION
Palo Alto‘s fire department has 7 engines (2 reserve), 1 type III brush truck which was
only staffed in the summer, 1 type III truck cross-staffed by Truck 6, 1 ladder truck, 3 ALS
ambulances (1 reserve), 1 rescue (1 reserve). These units are deployed in 8 stations. A total of 30
full time firefighters are staffed in 8 stations, and 4 administrative staff is staffed in station 6.
Two overtime paid firefighter paramedics are staffed on ALS ambulance M1 in station 1 from
0800 to 2000 hours. Hourly EMTs are staffed in BLS ambulance BLS1 in station 6 for 10 hours
a day, Monday through Friday. Between July and October, 3 additional fire fighters are staffed
for the brush truck in station 8 between 0800 and 2000 hours.
Our data analysis is divided into three sections. The first section focuses on the call types
and dispatches. The second section explores time spent and workload of individual units. The
third section presents response time analysis.
The data in this report cover all calls for service between July 1, 2009 and June 30, 2010.
During this period, Palo Alto‘s fire department received 6,752 non-canceled calls. Multiple units
are often sent to calls. A total of 15,625 Palo Alto units were dispatched during this period. The
total workload for the year for all Palo Alto units combined was 8,141 hours. Lastly, the average
total response time was 6.5 minutes for both EMS and fire category calls.
Calls per
Call Type Number of Calls Call Percentage
Day
BLS Transport 391 1.1 5.2
MVA 253 0.7 3.3
EMS Other 3,882 10.6 51.3
EMS Total 4,526 12.4 59.9
Structure Fire 49 0.1 0.6
Outside Fire 138 0.4 1.8
Hazard 272 0.7 3.6
Alarm Subtotal 1,104 3.0 14.6
False Alarm 169 0.5 2.2
Malfunction Alarm 382 1.0 5.1
Unintentional Alarm 458 1.3 6.1
Other Alarm 95 0.3 1.3
Public Service 389 1.1 5.1
Good Intent 274 0.7 3.6
Fire Total 2,226 6.1 29.4
Canceled 808 2.2 10.7
Total 7,560 20.7 100
Note: In the table Alarm Subtotal is further broken down into four categories: False, Malfunction,
Unintentional and Other.
Observations:
A total of 2.2 percent of calls were canceled.
On average, the department received 18.5 non-canceled calls per day.
EMS calls for the year totaled 4,526 (59.9 percent of all calls), or about 12.4 per day.
Number of Calls
Call Type Automatic aid Automatic aid Mutual aid Mutual aid
Total
given received given received
BLS Transport 16 16
MVA 5 61 1 67
EMS Other 12 269 281
EMS Total 17 346 1 364
Structure Fire 7 5 7 19
Outside Fire 3 15 2 20
Hazard 2 12 14
Alarm 7 8 15
Public Service 10 1 11
Good Intent 16 16
Fire Total 19 66 3 7 95
Canceled 78 58 1 137
Total 114 470 4 8 596
Note: Auto Aid is a predetermined response into another jurisdiction when the responding units from one
agency are physically closer to the incident location than the closest unit from the agency where the incident is
actually located. Mutual aid is a request for specific pieces or types of equipment when the requesting agency does
not have the necessary resources to respond to an incident.
Observations:
A total of 470 calls have received automatic aid, among which 58 (12.3 percent) calls
were canceled.
Palo Alto has requested mutual aid assistance for 7 structure fire calls and 1 motor
vehicle accident calls.
Palo Alto fire department has provided automatic aid for 114 calls, among which 78 (68.4
percent) calls were canceled.
Palo Alto fire department has provided mutual aid assistance for 2 outside fire calls and 1
public service call.
.......
_2ho<.s (1 ,1D2)
~ 2 hout. (::110)
1·2-"'(22)
--
111~>
Observations:
A total of 187 structure fire and outside fire calls accounted for 8.4 percent of fire
category total.
Alarm calls were 49.6 percent of fire category calls.
Public service calls were 17.5 percent of fire category total.
Hazardous condition calls were 12.2 percent of fire category total.
Good intent calls were 12.3 percent of fire category total.
Basic life support and transport calls accounted for 8.6 percent of EMS category total.
Motor vehicle accident calls were 5.6 percent of EMS category total.
Other EMS calls were 85.8 percent of EMS category total.
Observations:
Average calls per day varied by month and ranged from a low of 16.8 calls per day in
November 2009 to a high of 20.3 calls per day in February 2010, or 21 percent more than
the lowest month.
Average EMS calls per day varied from a low of 11.2 calls per day in November 2009 to
a high in 14 calls per day in February 2010.
Average fire category calls per day varied from a low of 4.8 to a high of 7.0 calls per day
(lowest in March 2010, highest in January 2010).
Unit
Call Type Five or
One Two Three Four Total
more
BLS Transport 210 13 136 29 3 391
MVA 8 41 94 63 47 253
EMS Other 389 2,814 578 79 22 3,882
EMS Total 607 2,868 808 171 72 4,526
Structure Fire 12 5 2 3 27 49
Outside Fire 87 18 1 1 31 138
Hazard 112 63 6 15 76 272
Alarm 896 71 6 25 106 1,104
Public Service 305 59 22 3 389
Good Intent 77 137 36 10 14 274
Fire Total 1,489 353 73 54 257 2,226
Grand Total 2,096 3,221 881 225 329 6,752
Percentage 31.0 47.7 13.0 3.3 4.9 100
Observations:
Total deployed time for the year, or total busy hours, was 8,141 hours. This is the total
deployment time of all the units that were deployed on any type of call.
There were a total of 15,625 runs, an average of 43 runs per day.
Fire category calls accounted for 17 percent of the total workload.
There were a total of 511 runs for structure and outside fire calls with a total workload of
337 hours, which accounted for 4.1 percent of the total workload. The average busy time
for structure fire calls was nearly one hour per call and for outside fire calls it was 26
minutes.
Alarm calls accounted for 5.5 percent of the total workload. The average busy time for
alarm calls was 15.6 minutes per call.
EMS calls accounted for 81.1 percent of the total workload. The average busy time for
EMS calls was 40.2 minutes per call.
Observations:
Engine Company E1 made 1,805 runs in a year, averaging 4.9 runs and nearly 2 hours of
busy time per day.
Engine Company E2 made 895 runs in a year, averaging 2.4 runs and 55 minutes of busy
time per day.
Engine Company E3 made 1,110 runs during the year, averaging 3 runs and one hour and
13 minutes of busy time per day.
Engine Company E4 made 1,016 runs during the year, averaging 2.8 runs and one hour
and 15 minutes of busy time per day.
Engine Company E5 made 816 runs during the year, averaging 2.2 runs and one hour of
busy time per day.
Engine Company E6 made 1,366 runs during the year, averaging 3.7 runs and one hour
and 29 minutes of busy time per day.
Engine Company E7 made 127 runs and were busy 41 hours in a year.
Brush engine E8B was only staffed in July, August, September 2009 and June 2010.
Engine E8B made 30 runs and were busy 17 hours.
Ladder truck T6 made 519 runs during the year, averaging 1.4 runs and 28 minutes of
busy time per day.
ALS ambulance M1 made 1,493 runs during the year, averaging 4.1 runs and 3 hours 31
minutes of busy time per day.
ALS ambulance M3 made 310 runs during the year, averaging 0.8 runs and 33 minutes of
busy time per day.
ALS ambulance M2 made 2,850 runs during the year, averaging 7.8 runs and 6 hours 34
minutes of busy time per day.
Rescue unit R2 made 856 runs during the year, averaging 2.3 runs and 48 minutes of
busy time per day.
Two-Hour
EMS Fire Total
Interval
0-1 28.2 4.9 33.1
2-3 19.2 4.3 23.5
4-5 16.3 5.7 22.0
6-7 31.9 6.1 38.0
8-9 64.2 14.3 78.5
10-11 74.2 11.0 85.3
12-13 75.8 13.3 89.1
14-15 67.1 14.6 81.6
16-17 54.3 10.6 64.9
18-19 45.2 13.5 58.6
20-21 37.6 7.1 44.8
22-23 28.9 8.1 37.0
Daily Total 1085.9 227.1 1313.0
Note: Daily total will equal to the sum of the column multiplied twice since each row represents two hours.
Observations:
The total busy minutes are the sum of both EMS and fire call workloads for all units.
Total busy minutes by hour varied significantly mainly due to the fluctuations of EMS
workload.
Runs
Structure Outside Public Good Total
Unit EMS Hazard Alarm per
Fire Fire Service Intent Runs
Day
E
1 1,311 20 34 59 213 95 73 1,805 4.9
E 5
2 13 19 26 70 188 36 43 895 2.4
E 6
3 78 21 28 76 136 110 61 1,110 3.0
E 6
4 70 13 33 60 121 84 35 1,016 2.8
E 5
5 22 11 24 47 129 36 47 816 2.2
E 7
6 41 21 56 61 415 30 42 1,366 3.7
E 2
7 0 2 5 11 85 1 3 127 0.3
T 1
6 73 32 31 90 134 31 28 519 1.4
E 1
8B 6 6 5 1 2 30 0.2
Note: Engine E8B is only staffed between July and October, so the daily average is calculated using total
days of 123.
Observations:
Engines E1, E3, E4 and E6 were each deployed more than 1,000 times in a year.
Engine E1 responded to the most EMS calls, averaging 3.6 EMS calls per day.
Observations:
ALS ambulance units M1 and M2 were deployed 1,493 and 2,850 runs, respectively, in a
year, averaging 4.1 and 7.8 runs per day.
On average, ALS ambulance unit M3 was deployed less than 1 run per day.
On average, rescue unit R2 wad deployed 2.3 runs per day.
Observations:
Except rescue unit R2, EMS calls accounted for more than 96.8 percent of total workload
for all ALS and BLS ambulance units.
On average, ALS ambulance unit M1 was busy for 3 hours and 31 minutes per day.
On average, ALS ambulance unit M2 was busy for 5 hours and 44 minutes per day.
On average, BLS ambulance unit BLS1 was busy for 2 hours and 1 minute per day.
On average, rescue unit R2 was busy for 48 minute per day, and EMS calls accounted for
46.5 percent of the total.
We use different terms to describe the components of response time. Dispatch processing
time is the difference between the unit dispatch time and the call receipt time. Turnout time is
the difference between the unit time en route and the unit dispatch time, while travel time is the
difference between the unit on-scene arrival time and the unit time en route. Response time
includes dispatch processing time, turnout time, and travel time.
Figure 7. Average Dispatch, Turnout, Travel, and Response Time of First Arriving Units
by Call Type
90th
Dispatch Turnout Travel Percentile Number
Call Type Response Time
Time Time Time Response of Calls
Time
BLS Transport 6.5 2.0 9.0 16.7 43.7 356
MVA 0.5 1.9 4.0 6.3 9.9 235
EMS Other 0.4 1.8 3.5 5.6 7.7 3,829
EMS Total 0.9 1.8 4.0 6.5 8.3 4,420
Structure Fire 0.5 2.5 3.4 6.2 7.9 42
Outside Fire 0.5 2.1 4.3 6.7 8.9 134
Hazard 0.6 2.0 4.2 6.8 9.8 265
Alarm 0.4 2.1 3.9 6.2 8.9 1,013
Public Service 0.5 1.9 5.1 7.4 11.4 382
Good Intent 0.5 1.8 3.6 5.8 8.3 265
Fire Total 0.5 2.0 4.1 6.5 9.4 2,101
Total 0.8 1.9 4.1 6.5 8.9 6,521
Note: The average of dispatch time, turnout time, travel time and response time are calculated using non-
null values of each field and they may have different number of calls. Therefore, the average response time may not
be equal to the sum of average dispatch time, turnout time and travel time.
Observations:
The average dispatch time for all calls was 0.8 minutes. However, the average dispatch
time for BLS transport calls was 6.5 minutes.
The average turnout time was 1.9 minute.
The average travel time was 4.1 minutes. However, the average travel time for BLS
transport calls was 9 minutes.
The average response time for all EMS calls was 6.5 minutes and the 90th percentile
response time for all EMS calls was 8.3 minutes.
The average response time for fire category calls was 6.5 minutes and the 90th percentile
response time for fire category calls was 9.4 minutes.
The average response time for structure fire calls was 7.9 minutes.
The average response time for outside fire calls was 8.9 minutes.
Table 13. Number of Total Calls for the First Arriving Units
6
.
• 5
;
o
~ 4
3
,
]
o
o 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23
Hour 01Oil Y
Observations:
Average dispatch time was between 0.1 and 1.5 minutes.
Average turnout time was between 1.5 and 2.7 minutes. Between 9 a.m. and 11 p.m.,
average turnout time was less than 2 minutes.
Average travel time was between 3.4 and 5.1 minutes.
Average response time was between 5.5 and 8.1 minutes.
The vertical axis is the probability or percentage of calls. The horizontal axis is response
time. For example, with regard to MVA and EMS other calls, the 0.9 probability line
intersects the graph at the time mark at about 7.8 minutes. This means that units
responded to 90 percent of these calls in fewer than 7.8 minutes.
Response Cumulative
Frequency
Time Percentage
1 min. 44 1.1
2 min. 61 2.6
3 min. 128 5.7
4 min. 442 16.6
5 min. 942 39.8
6 min. 1,084 66.5
7 min. 648 82.4
8 min. 373 91.6
9 min. 133 94.9
10 min. 70 96.6
11 min. 42 97.6
12 min. 28 98.3
13 min. 13 98.6
14 min. 8 98.8
15 min. 5 98.9
>= 15 min 43 100.0
Note: BLS transport calls have significantly longer response time and therefore not included.
Observations:
The average response time for MVA and EMS other calls was 5.7 minutes.
For 66.5 percent of EMS calls, the response time was less than 6 minutes.
For 90 percent of EMS calls, the response time was less than 7.8 minutes.
The following tables and charts report our response time analysis of first arriving units
for structure and outside fire calls. The analysis focuses on the arrival of firefighting equipment
including engine, and truck units. The response time analysis does NOT include dispatched
ambulance units for structure and outside fire calls. They typically arrive along with the Engine
Company based in their station.
Table 16. Average Response Time for Structure Fire and Outside Fire Calls by First
Arriving Fire Unit
Observations:
For outside fire calls, Engine E1 and E4 were the first unit on scene most often and had
the shortest average response time of 5.4 and 5.2 minutes respectively when arriving first
on scene.
For structure fire calls, Engine E1 was the first unit on scene most often.
Except two runs of E8B, first arriving units arrived on scene on average in fewer than 10
minutes.
Observations:
Ladder truck T6 and Engine E3 were the second unit on scene most often.
When E6 were the second unit on scene, it had the shortest response time of 6.8 minutes.
All second arriving units arrived on scene on average in fewer than 14 minutes.
100
80
2nd unit average: 9.1 minutes /' ,/
/ ~ ~
3rd unit average: 10.2 minutes
•
g
70
/ / /
60
,
~
~
• 50 / / /
/ / /
.,~E 40
/ / /
, 30 1st unit 90th percentile: 8.7 minutes
u
20 / / / 2nd unit 90th percentile: 16.7 minutes
3rd unit 90th percentile: 14.2minutes
10 / / ./
0
./ /~
o I 4 6 7 8 9 10 11 12 13 14 15
,~-,C'-'CUc,c,;C,~---:,-"CdCUC,C,;C'~---:3-nCICuC"C"~1 Minutes
Structure Fire
First Unit Second Unit Third Unit
Response Time
Cumulative Cumulative Cumulative
Frequency Frequency Frequency
Percent Percent Percent
1 min. 3 1.7 0.0 0.0
2 min. 3 3.4 0.0 0.0
3 min. 3 5.1 0.0 0.0
4 min. 19 15.9 0.0 0.0
5 min. 39 38.1 1 2.0 0.0
6 min. 34 57.4 5 11.8 1 3.1
7 min. 28 73.3 12 35.3 3 12.5
8 min. 21 85.2 10 54.9 4 25.0
9 min. 12 92.0 9 72.5 6 43.8
10 min. 3 93.8 2 76.5 7 65.6
11 min. 2 94.9 4 84.3 1 68.8
12 min. 1 95.5 1 86.3 1 71.9
13 min. 1 96.0 1 88.2 2 78.1
14 min. 2 97.2 0 88.2 2 84.4
15 min. 1 97.7 0 88.2 2 90.6
16 min. 0 97.7 0 88.2 1 93.8
17 min. 2 98.9 2 92.2 1 96.9
18 min. 0 98.9 1 94.1 0 96.9
19 min. 1 99.4 0 94.1 0 96.9
20 min. 0 99.4 3 100.0 1 100.0
20 to 25 min. 1 100.0 0 100.0 0 100.0
Observations:
The average response time of the first arriving fire unit for structure and outside fire calls
was 6.6 minutes.
90 percent of the time, the first fire unit arrived on scene in fewer than 8.7 minutes.
On average, the response time of the second arriving unit was 9.1 minutes, which was 2.5
minutes longer than that of the first arriving unit.
90 percent of the time, the second fire unit arrived on scene in fewer than 16.7 minutes.
The hour with the most calls received was between 0800 and 0900 on February 17, 2010.
The 11 calls involved 12 runs. The combined workload was 326 minutes, which may spread
more than one hour. This is equivalent to between 5 and 6 firefighting units being busy the entire
hour. However, in the city there are 13 units staffed all of the time. During the worst portion of
the hour, there were always at least 6 units still available to respond immediately. Only 7 of the
13 units were busy more than 30 minutes during this hour.
The hour with the most calls was between 2100 and 2200 on September 19, 2009. The 8
calls involved 11 runs. The combined workload was 252 minutes, which may spread more than
one hour. This is equivalent to between 4 and 5 firefighting units being busy the entire hour.
However, in the city there are 14 units staffed all of the time in the summer. During the worst
portion of the hour, there were always at least 8 units still available to respond immediately.
Only 1 of the 14 units was busy more than 30 minutes during this hour.
Number of Calls in
Frequency Percentage
an Hour
0 4281 48.9%
1 2865 32.7%
2 1122 12.8%
3 373 4.3%
4 86 1.0%
5 26 0.3%
6-10 6 0.1%
11-15 1 0.0%
Observations:
A total of 33 hours (0.4 percent) in a year, there are 5 or more calls occurred in an hour.
It is approximately once every 11 days the fire department will respond to 5 or more calls
in an hour.
Observations:
The hour with the most calls received was between 0800 and 0900 on February 17, 2010.
The 11 calls involved 12 runs. The combined workload was 326 minutes, which may
spread more than one hour. This is equivalent to between 5 and 6 firefighting units being
busy the entire hour. However, in the city there are 13 units staffed all of the time.
Station 1 2 3 4 5 6 7
Number of
E M M E M R E E E B E T E Units
Busy
Unit
1 1 3 2 2 2 3 4 5 LS1 6 6 7
0-5 5.0 0.6 5.0 5.0 4
5-10 0.3 2.2 1.7 5.0 5.0 5.0 6
10-15 3.6 1.6 5.0 3.9 4.9 5.0 5.0 7
15-20 4.8 3.7 0.2 4.1 0.7 5
20-25 5.0 5.0 5.0 2.7 3.7 5
25-30 5.0 5.0 5.0 5.0 0.7 5.0 6
30-35 5.0 5.0 5.0 0.8 5.0 5.0 5.0 7
35-40 5.0 5.0 5.0 5.0 5.0 5.0 5.0 7
40-45 5.0 5.0 5.0 5.0 5.0 5.0 4.7 7
45-50 3.4 5.0 4.2 5.0 5.0 5.0 5.0 7
50-55 0.2 5.0 5.0 5.0 5.0 5
55-60 5.0 2.5 5.0 5.0 4
Total 42.1 1.6 0.0 41.1 0.0 0.0 35.0 36.3 54.3 30.7 54.1 0.0 0.0
Note: The numbers in the cells are the busy minutes within the 5 minute block. The cell values greater than 2.5 are coded as red.
Observations:
During the worst portion of the hour, there were always at least 6 units still available to respond immediately. Only 7 of the 13
units were busy more than 30 minutes during this hour.
Observations:
The total busy minutes for EMS calls were 228.6 minutes, which accounted for 77.4
percent of the total buys minutes.
The total busy minutes for non structure and outside fire category calls were 66.6
minutes, which accounted for 22.6 percent of the total buys minutes.
Station 1 2 3 4 5 6 7 8 Number of
Unit E1 M1 M3 E2 M2 R2 E3 E4 E5 BLS1 E6 T6 E7 E8B Busy Units
0-5 0
5-10 0
10-15 0
15-20 3.0 1.1 1.1 3
20-25 5.0 5.0 5.0 3
25-30 5.0 5.0 3.3 2.7 3.3 5.0 6
30-35 5.0 4.7 5.0 5.0 5.0 4.5 6
35-40 5.0 5.0 4.6 4.3 4
40-45 5.0 5.0 3.7 2.5 4
45-50 5.0 5.0 5.0 5.0 4
50-55 5.0 1.2 2.7 2.9 2.8 1.2 6
55-60 2.9 5.0 5.0 3
Total 40.9 0.0 0.0 0.0 22.0 26.0 0.0 23.9 22.9 21.8 0.0 0.0 0.0 0.0
Note: The numbers in the cells are the busy minutes within the 5 minute block. The cell values greater than 2.5 are coded as red.
Observations:
During the worst portion of the hour, there were always at least 8 units still available to respond immediately since E8B was
staffed. Only 1 of the 13 units were busy more than 30 minutes during this hour.
Observations:
The total busy minutes for EMS calls were 43.8 minutes, which accounted for 27.8
percent of the total buys minutes.
The total busy minutes for non structure and outside fire category calls were 113.7
minutes, which accounted for 72.2 percent of the total buys minutes.
NFIRS
Incident Description Call Type
Incident Code
321B BLS Interfacility Transport Only BLS Transport
321D EMS call, downgraded for BLS transport BLS Transport
Vehicle Accident with Injuries - Downgraded for
322D BLS Transport
BLS Transport
323D Motor vehicle/pedestrian accident BLS downgrade BLS Transport
322 Vehicle accident with injuries MVA
323 Motor vehicle/pedestrian accident (MV Ped) MVA
324 Motor vehicle accident with no injuries MVA
321 EMS call, excluding vehicle accident EMS Other
322 Vehicle accident with injuries EMS Other
352 Extrication of victim(s) from vehicle EMS Other
353 Removal of victim(s) from stalled elevator EMS Other
357 Extrication of victim(s) from machinery EMS Other
445 Arcing, shorted electrical equipment EMS Other
463 Vehicle accident, general cleanup EMS Other
510 Person in distress, other EMS Other
600 Good intent call, other EMS Other
661 EMS call, party transported by non-fire agency EMS Other
321P EMS call, 5150 Transport EMS Other
111 Building fire Structure Fire
112 Fires in structures other than in a building Structure Fire
113 Cooking fire, confined to container Structure Fire
114 Chimney or flue fire, confined to chimney or flue Structure Fire
130 Mobile property (vehicle) fire, other Structure Fire
461 Building or structure weakened or collapsed Structure Fire
100 Fire, other Outside Fire
118 Trash or rubbish fire, contained Outside Fire
131 Passenger vehicle fire Outside Fire
140 Natural vegetation fire, other Outside Fire
141 Forest, woods or wildland fire Outside Fire
Total
Agency Unit Description Unit Number Annual
of Calls Busy Hours
A2 1 0.6
Deputy Chief
A3 1 2
Battalion Chief B6 500 148.5
Palo Alto
Training Captain TC1 2 1.5
K1 4 14.6
Fire Investigator
K2 1 2
COE14 4 1.4
County Fire Engine COE15 19 3.4
COE16 1 0.1
County Fire Truck COT14 1 1.6
AMR1 531 159.8
AMR2 74 27
County ALS Ambulance AMR3 58 17.3
AMR4 8 1.8
AMR5 1 0.2
County Fire Rescue COR14 10 2.1
COB7 7 2.6
County Fire Bat Chief
COBC 1 1.1
MPE1 54 13.1
Menlo Park Engine MPE4 6 1.1
Menlo Park
MPE6 44 2.6
Menlo Park Truck MPT1 1 0
MVE3 40 7.6
Mountain View Engine
MVE5 51 9
Mountain
Mountain View Truck MVT1 11 9.2
View
Mountain View Rescue MVR1 9 1.5
Mountain View Bat Chief MVB18 10 7.1
FHP1 8 6.7
Park Ranger
FHP2 13 10.6
The PAFD provides service from seven stations year-round and one station on a seasonal
basis. The seasonal station (8) is staffed for 90 days (July through October), which is considered
the high-fire season. Palo Alto has an (ISO) Insurance Services Office rating of two, with the
highest being one. This rating is a technical guide for determining insurance costs and not
intended to be a guide in the design of a fire department. This is especially true for fire
departments that provide emergency medical service since EMS is not considered within the
rating schedule.
In this Appendix we discuss some additional findings related to improving the overall
management of the PAFD. We also present a plan to improve accountability, internal discipline,
and empower employees to effectively embrace change through project management. Also
included is a template the fire department to consider for its project management system.
New habits and skills will have to be developed, as well as integrating the initiatives and
evaluative components of this plan at all levels of the department. Consequently, we recommend
installing the AIM system, which would serve as the foundation for establishing new individual
and organizational practices. AIM—an acronym for Assessment, Improvement, Maintenance—is
designed to enhance individual, team, operational, administrative, and interpersonal functions by
supporting strong communication practices, continuous quality improvement, and leadership
development. AIM is a FIRE/EMS performance management program offered by Performance
Management Group, San Rafael, California. A minimum review of quarterly review and
discussion of related issues by management personnel is recommended at staff meetings with
provision for feed-back of answers or resulting direction.
2. Revise performance evaluations to achieve organizational goals. A key need for the
Palo Alto Fire Department is to have supervisors build the confidence necessary to evaluate the
Step 1: Establish a Planning Team and Gather Information – The first step is to
create a planning team to focus on the broad spectrum of areas associated with
planning such as the implementation of the recommendations in this report and an
annual review of the Departments strategic plan.
The team should be small enough to facilitate good group discussions, but large
enough to include appropriate goal setters, analysts for measuring performance
against metrics, and administration responsible for implementing the annual strategic
plan. The following are a good starting point for individuals to include: chief; chief
officer; fire captain; fire fighter; medical director; medical coordinator; fire marshal;
public education person; IT Specialist; GIS analyst; labor representative and business
manager.
Performance Measurements
The reports generated by the operations and data analysis team are based on key performance
indicators that have been identified in standards and safety regulations and by special interest
groups such as the International Association of Chiefs of Police, the International Association of
Fire Chiefs, the International Association of Fire Fighters, and the Association of Public Safety
Communication Officials International as well as and through the Center for Performance
Measurement of ICMA. These performance indicators have developed from decades of research
and are applicable in all communities. For that reason, comparisons of reports reveal similar
reporting formats, but each community‘s data are analyzed on an individual basis by the ICMA
specialists and uniquely represent the compiled information for that community.
We begin most projects by extracting calls for service and raw data from an agency‘s computer-
aided dispatch system. The data are then sorted and analyzed for comparison with nationally
developed performance indicators (e.g., response times, workload by time, multiple unit
dispatching), which are valuable measures of agency performance regardless of departmental
size. The findings are shown in tabular as well as graphic form. Our documents generally follow
a standard process and rely on standard categories for performance indicators; because of the size
and complexity of the documents, this allows for simple, clean reporting. Again, however, the
data reported are unique to the city that is the subject of a particular report.
In many cases, cities hire ICMA to conduct an operational review alongside a data analysis. In
those cases, the performance indicators serve as the basis for the operational review and so the
review process also follows a standardized approach, one that is comparable to that of national
accreditation agencies. Before an ICMA Center for Public Safety Management team arrives on-
site, agencies are asked to compile a number of key operational documents (policies and
procedures, asset lists, etc.). Most on-site reviews consist of interviews with management and
supervisors as well as with rank-and-file officers and firefighters; attendance at roll calls; and
meetings with elected officials. We review call records and observe dispatch operations to ensure
compliance with the provided written documentation.
From any on-site visits and data assessments that have been conducted, our subject matter
experts produce observations and recommendations that highlight the department‘s strengths,
weaknesses, opportunities, and threats. We have found that this standardized approach ensures
that we measure and observe all critical components of a public service department, which in
We liken this standardized approach to the manner of the financial audit: ICMA asks questions and
requests documentation upon project start-up; confirms the accuracy of information received;
deploys operations teams on-site to research the uniqueness of each environment; performs data
modeling and shares preliminary findings with each city; assesses any inconsistencies reported by
client cities, and, finally, communicates our results in a formal, written report and occasionally
with an in-person presentation hosted by the project team and other key contributors.
Our approach has been tried and tested, and we are confident that as public safety issues and best
practices evolve as a result of enhanced technology and intelligence and as cities‘ profiles change
because of population shifts or other key factors, our approach will remain relevant, effective, and
highly successful.
A. Administration
The primary challenge facing the department is the establishment of an effective and
sufficient command structure. Because of retirements and the lack of succession planning, the
administration of the department is overwhelmed. As a result, it has resorted to managing by
tactic versus by an agreed upon strategy with corresponding tactics provided to achieve the
strategic goals.
ICMA and TriData both agree that the current arrangement of overseeing the department
by the police chief should be strengthened by creating a ―
Director of Public Safety Services.‖
We came to that conclusion after meetings with citizens, elected and appointed officials, and
departmental staff. The police chief is currently filling this role and both firms feel that
recognizing the value will immediately create additional depth in the administration for the fire
department.
Serving the director would be a position like that of a fire chief and police chief.
By creating a ―
Director of Public Safety,‖ one other issue can be immediately resolved:
the Emergency Management Department. Those functions are currently shared between police
and fire with responsibilities in each and without a clear delineation of function should an
emergency occur. By reporting to one person – the director – the emergency management
functions can be better coordinated and a more administration developed.
Palo Alto has an outstanding Citizen Emergency Response Team (CERT) with volunteers
throughout the community. By creating a structure under the director and utilizing both police
and fire department resources, the capacity of this unit can be fully realized. Because it is
volunteer, it is also cost effective for the city to build the CERT rather than hiring full time
personnel.
C. Facilities
There is an adage that ―afire facility is a tool, not just a shelter‖. The design of a fire
facility greatly affects the manner in which an agency operates. The facility provides the right
mix of security and accessibility.
E. Information Technology
The completion of the TriData work was hindered by the inability to analyze records of
the department.
Both ICMA and TriData have found that fire department record systems are not as
vigorous as their police counterparts. Police departments must capture and report data accurately
or face legal sanctions and dismissal of prosecution. Fire departments, on the other hand, do not
face sanctions for errors in reporting or systems that do not capture accurate data.
In order to strategically manage the department, accurate data on performance must be
captured and be capable of analysis. This analysis should not be like that done by these reports; it
needs to be done on a monthly, weekly, or even daily basis. An agency or city cannot improve on
something for which they do not have a measurement process.
F. Dynamic Deployment
EMS consumes the majority of time for the Palo Alto Fire Department. Time to reach a
patient is critical to recovery or successful intervention. The analysis of station locations and
travel time conducted by TriData clearly showed the affects of traffic calming and the street
network on the efficiency and effectiveness of deployed resources. The data analysis conducted
by ICMA also showed that first unit response at the 90th percentile was within the range of
national benchmarks established by groups such as the National Fire Protection Association
(NFPA) in its 1710 deployment standard. However, second and third units were higher than the
recommended ranges.
In order to begin patient care in under 10 minutes from the initial call for service or to
deploy sufficient resources to allow two firefighters to enter a burning structure (using the two-
in; two-out rule of OSHA), 9.1 minutes of time is consumed. This is shown in Figures 10 and 11
of the data analysis portion of this report.
The alternatives available to correct this problem include: remove traffic calming, build
more stations, deploy with a dynamic model.
The first two alternatives would not only be costly, they would likely be met with
opposition from neighborhood associations. The third alternative – dynamic deployment – uses
data to determine where calls for service are likely to be generated. Research has shown that
human beings and communities are quite predictable; calls for service create trends and patterns
that can be used predictably to locate resources in the event responders are required. The
research has been intensively developed in the United Kingdom over the past nine years to also
identify preventive mechanisms that can be taken to eliminate calls altogether.
Deploying dynamically would position resources using computer models that would
utilize data collected from department records. Because the data is currently not in a form to be
easily analyzed, one critical component to success of this strategy would be the creation of a
Geographic Information System (GIS)/Technology position. Again, this could be shared with the
police department and other city services.
H. Alarms
False alarm issues are faced by all departments; ICMA has found college towns experience
higher rates than average.
The data analysis showed Palo Alto responded to 1,714 alarms which consumed an
average of 15.6 minutes per response. The total spent on answering alarms was 445 annual and
was responsible for 5.5% of the available time. Alarm response was second only to EMS calls
for service.
ICMA has developed strategic relationships with a number of firms that are applying
technology to eliminate one of the leading calls for alarms: smoke from stoves and microwaves.
Pioneering Technology has developed a replacement burner for electric stoves and hot plates that
reaches a maximum heat temperature sufficient to cook but that minimizes fires. Normal electric
ranges and hot plates generate substantially more heat than is required to cook (and consume
considerably more energy as a result). The technology developed replaces the burners with
―
smart technology‖ that eliminates this danger and saves substantial energy in the process. It has
12
Leonard Matarese, David Limardi, “The Value of Automated Defibrillators Lives On,”
Public Management 92 (September 2010 p. 15-17)