software documentation

Transcription

Draft version: 2012-02-09
www.baseform.org/np4/awareApp
AWARE-P
© 2012
Diogo Vitorino
Sergio T Coelho
Helena Alegre
André Martins
João Paulo Leitão
Maria Santos Silva
Draft version: 2012-02-09
Acknowledgements
The AWARE-P software is developed in the context of the AWARE-P project, a leading edge R&D
effort funded by the European Economic Area (Contract No. PT 0043) and by the project partners.
The project was developed and co-funded by a consortium led by LNEC – National Civil
Engineering Laboratory (Portugal) and comprising IST – Instituto Superior Técnico
(Technical University of Lisbon, Portugal), SINTEF (Norway), Addition (Portugal) and
YDreams (Portugal), as well as by ERSAR – Water and Waste Services Regulator (Portugal),
and by the AWARE-P end-user partners: AGS S.A., AdP Serviços S.A, SMAS Oeiras &
Amadora and Veolia Águas de Mafra.
Software development benefited from suggestions and contributions from a large number
of team members and project friends, including: Adriana Cardoso, André Martins, André
Pina, Daniel Mendes, Didia Covas, Diogo Vitorino, Enrique Cabrera, Helena Alegre, João
Feliciano, João Paulo Leitão, Joaquim Beleza, Julieta Marques, Kjersti Holte, Luís Loureiro,
Luís Mamouros, Maria do Céu Almeida, Maria Santos Silva, Nelson Carriço, Pedro
Ramalho, Pedro Rufino, Pedro Pereira, Rita Ugarelli, Rodrigo Borba, Rui Rua, Sérgio T
Coelho and Sigurd Hafskjold.
License
The software described in this document is distributed under the GNU General Public
License. For further details go to: www.baseform.org/np4/aboutLicense.html
For information on the several components used, see the “Detailed Licensing” section in:
Disclaimer
Although all efforts have been undertaken to ensure that the software described here is
of the highest possible quality and that the results obtained are correct, the authors do
not warrant the functions contained in the program will meet your requirements or that
the operation of the program will be uninterrupted or error-free. The authors are not
responsible and assume no liability for any results or any use made thereof, nor for any
damages or litigation that may result from the use of the software for any purpose.
—
Google, Chrome, Google Earth, Mozilla, WebGL, Firefox, Apple, Mac, Safari, Windows,
Microsoft, Word, Excel, Bing, OpenStreetMaps and all other trademarks and copyrights
mentioned herein are the property of their respective owners.
Contents
7 AWARE-P
7
7
8
12
12
13
Purpose
Overview
Details
Usage
Further reading
See also
15 TOOLS
15
15
15
15
16
16
PLAN – AWARE-P planning
Purpose
Overview
Usage
Further reading
See also
17
17
17
17
17
19
NETWORK – EPANET
Purpose
Overview
Details
Usage
Further reading
20
20
20
20
21
21
22
PI – PERFORMANCE INDICATORS
Purpose
Overview
Details
Usage
Further reading
See also
23
23
23
23
24
24
24
PX – Performance Indices
Purpose
Overview
Details
Usage
Further reading
See also
25
25
25
26
27
27
FAIL – Failure Analysis
Overview
Details
Usage
Further reading
See also
27
27
27
28
28
28
28
CIMP – Component importance
Purpose
Overview
Details
Usage
Further reading
See also
29
29
29
29
29
29
UNMET – Expected Unmet Demand
Purpose
Overview
Details
Usage
See also
30
30
30
30
31
31
31
IVI – Value Index
Purpose
Overview
Details
Usage
Further reading
See also
32 CORE
32 Purpose
32 Overview
32 Details
34 Appendix A
AWARE-P software documentation: draft version 2012-02-02
p. 4 / 38
FOREWORD
This document aims at helping users understand the purpose of the AWARE-P planning software and provides guidance on how to use it. The document describes the application in general terms, before introducing the tools that are included, and the core software platform where
they exist. Each section is organized into purpose, overview, details, usage and further reading.
The software is publicly available at www.baseform.org, and can be accessed through a simple free registration procedure. It runs from any common browser (for best results, use Google
Chrome®, Mozilla Firefox® or Apple Safari®, on any Windows, Mac or Linux system).
Further information, instructional materials and videos are available at www.baseform.org.
p. 5 / 38
p. 6 / 38
AWARE-PP
Purpose
The AWARE-P infrastructure asset management (IAM) planning software for drinking
water, wastewater and storm water services is an organized assessment environment where
planning alternatives or competing projects are measured up and compared through selected
performance, risk and cost metrics. It comprises a portfolio of metrics and analysis tools that
may be used individually for diagnosis and sensitivity gain purposes, or as part of the integrated planning procedure laid out by the AWARE-P IAM programme.
Overview
The infrastructure asset management approach developed in the AWARE-P project (www.
aware-p.org) is a broad management and engineering process aiming at alignment of objectives and targets, as well as effective feedback across the various decisional levels – strategic,
tactical, operational (Alegre et al., 2011).
The IAM process is fundamentally led by the stated objectives – and by an educated choice of
assessment criteria, well-chosen metrics and quantifiable targets. Based on the simple notion
that every investment in a system or any change to the way it is managed will most probably
impact not just one, but all three of the dimensions involved – performance, risk and cost –
the AWARE-P approach provides an unbiased and quantified framework for organizing the
task of generating, comparing and selecting alternatives for system improvement.
The AWARE-P IAM planning software makes available a coherent set of user-configurable
assessment models related to performance, cost and risk, which are used to evaluate userdefined alternative system modifications, planning solutions or competing projects, over a
given analysis period. Based on given planning objectives and measuring criteria, the user
p. 7 / 38
selects a set of metrics from the software’s metrics portfolio and proceeds to evaluate each
planning alternative at the selected time frames within the planning and analysis horizons,
feeding a cubic space of planning results.
The software’s tools may also be used in stand-alone, direct assessment mode for the fastest
possible path to results, or in the context of general-purpose sensitivity gain and system diagnosis. Examples of such uses may be: an analysis of failures rates (Poisson and LEYP models are available) and of risk of service interruption; a PI calculation (AWARE-P includes a
full-fledged PI tool with the most up-to-date libraries); a water quality simulation, exploring
the impact of alternative sites for a new rechlorination facility; or a fully hydraulic-enabled
investigation of network component importance (aka ‘criticality’).
The software’s tools have been specifically developed to make the best available methods and
analysis algorithms accessible for effective industry usage, retaining a maximum of simplicity in delivering meaningful and useable results.
The AWARE-P software is a web-based application that may be run on public or private
server, or as a local, stand-alone deployment. It is implemented using the open-source Baseform development platform and materializes as a growing set of plug-in tool modules made
available on that platform, taking advantage of its user management, common data integration services and next-generation 2D/3D visualization capabilities.
Details
The AWARE-P software provides the means to visualize, diagnose and evaluate any given
water supply, wastewater or stormwater system, through a portfolio of performance, risk and
cost models, at both global and detail levels; and, if so desired, to compare a system with
any number of planning alternatives or competing projects using standardized methods that
facilitate choice and decision-making – both manually and with the assistance of decisionsupport tools – tested against current or projected scenarios.
AWARE-P has essentially two main usage modes:
(i) as a portfolio of assessment-oriented models and analysis tools that may be used (individually or in
combination) in order to diagnose and gain sensitivity to a system; or
(ii) supporting the AWARE-P IAM planning procedure through to the definition of a planning framework
(time horizon, metrics, alternatives) and by feeding the planning tool with metric values produced
using the tools available.
PLAN is the tool that embodies the central planning framework of the AWARE-P infrastructure asset management programme, where planning alternatives or competing projects are
measured up and compared through selected performance, risk and cost metrics, through
interactive numerical and 2D/3D graphical information display.
AWARE-P hosts a growing number of plug-in tools that are as effective at producing metrics
that feed PLAN, as they are tailored for stand-alone usage, as fully-fledged analysis algorithms and models. The range of metrics-producing tools that are currently available, and
whose details are described further along in this document, include:
p. 8 / 38
• PI – Performance Indicators, quantitative assessment of the efficiency or effectiveness of a system
through the calculation of performance indicators based on state-of-the-art, standardized PI libraries as well as user-developed or customized ones.
• PX – Performance Indices, technical performance metrics based on the values of certain features
or state variables of water supply and waste/stormwater networks. The indices measure performance concepts related to level-of-service, network effectiveness and efficiency.
• FAIL – using models such as Poisson and LEYP, prediction of future pipe or sewer failures for a given
network, e.g. in the context of estimating risk or cost metrics, based on an organized failure history
in the form of work orders and pipe data records.
• CIMP – calculates a component importance metric for each individual pipe in a network, based on
the impact of its failure on nodal consumption. The measure is computed based on the network’s
hydraulic model, using full simulation capabilities.
• UNMET – calculates a service interruption risk metric expressed as the expected volume of unmet
demand in a system over one year, given the expected number of outages for each pipe, the average downtime per pipe outage, and the component importance of each pipe, expressed in terms of
unmet demand; system pipes are ranked accordingly.
• IVI – Infrastructure Value Index, representing the ageing degree of an infrastructure, calculated
through the ratio between the current value and the replacement value of the infrastructure.
• NETWORK-EPANET – an efficient, Java-implemented Epanet simulation engine and natively integrated MSX library, for full-range hydraulic and water quality network simulation. It takes advantage
of Baseform Core’s NETWORK and its 2D / 3D network and results visualization.
p. 9 / 38
Systems, not just collections of assets
The NETWORK tool is of crucial importance in the sense that it relates to one very important feature of the software and of the AWARE-P IAM approach: its focus on evaluating
water networks as systems rather than as collections of independent assets. For this reason,
and whenever needed, the available range of assessment models and methods draws on the
capability to simulate system behavior, either in simplified terms of by drawing support from
network simulators such as Epanet.
The entire set of visualization and analysis tools is available for exploratory use without having to follow a predefined project-driven script. From this viewpoint, the software is akin to
a wide-ranging, system modeling software, suited to what-if and sensitivity analyses and to
general system modeling.
Planning approach: the AWARE-P IAM programme
Good IAM is about finding the best possible balance of performance, risk and costs over a
long-term planning horizon. There are many progression paths that improve an urban water
system’s service performance or help control risks such as interruption of supply or water
quality incidents, and there will be a certain combination of interventions that will maximize the benefit of a given amount of investment. It is vital that many diverse alternative
solutions to improving the system are explored and compared, on a quantifiable and standardized basis.
AWARE-P has defined both a language and
a complete IAM programme to achieve that
goal. The infrastructure asset management
approach developed in the AWARE-P project
is a broad management process that addresses the need for a plan-do-check-act (PDCA)
philosophy at the various decisional levels
in a utility – strategic, tactical, operational
– aiming at alignment of objectives, metrics
and targets, as well as solid feedback across
levels (Alegre et al., 2011). This concept permeates the planning processes at each of the
levels, through the PDCA-inspired loop illustrated here.
The IAM process is fundamentally led by the stated objectives, and by an educated choice
of assessment criteria, metrics and quantifiable targets. This is particularly evident at the
strategic and tactical levels, the latter being the prime field of application for the software
described here.
Producing the plan is a problem-driven process, with a strong emphasis on thorough diagnosis in order to identify and assess the system’s main issues and shortcomings, in view of the
set targets, and to help decide where and how to act. Diagnosing and assessing a water supply,
wastewater or stormwater system, over given time horizons (at least the planning horizon
p. 10 / 38
and a longer, impact-analysis horizon), draw from a large range of methods and models for
evaluating performance, risk and cost (Alegre & Covas, 2010; Almeida et al., 2011). For this
purpose, a portfolio of techniques was selected that range from system statistics to network
simulation models, to hydraulic and water quality performance, to component failure analysis and forecasting, to component importance and criticality, and to methods for estimating
tangible capital and running costs.
The planning process is illustrated in the
schematic in very simple terms. The drawing board on the right-hand side is initially
marked out by the green vertical lines, representing the metrics for the criteria chosen to
drive the analysis. A thorough diagnosis and
assessment of the current system according
to those metrics is carried out (represented
by the first blue horizontal at the top). The
planning board is then successively populated with planning alternatives (represented
by the subsequent blue lines). The intersections represent the assessment of each planning alternative for each metric, and the
purpose of the process is to fill out the table
to the extent possible.
A separate table is calculated for each relevant time frame of the planning and analysis horizons, effectively giving rise to a cube of results, such as made available by the software.
The criteria draw from the available analysis methods in the performance, risk or cost dimensions. Examples could be hydraulic performance related to minimum available pressure
(as given by a hydraulic model), risk of interruption of supply due to pipeline failure (e.g.,
calculated by combining forecast failure rates with component importance derived from network analysis), or the net present value of a given alternative. The metrics used to evaluate
these criteria tend to lead to standardized quantities, which are more easily compared together and thus facilitate decision-making. Illustrations of this methodology can be found in
Marques et al. (2011) and Alegre et al. (2011).
p. 11 / 38
Usage
AWARE-P is accessed as a regular website, using any common web browser,(1) by addressing
a public www link (public versions), a private or enterprise server, or a local (stand-alone)
installation server. Versions where user-management has been activated will require a log in.
A general introduction to the principles and usage concepts behind the Core platform, where
the application is based, is given in the CORE section.
The software’s environment includes a main menu on the left side, and an expandable main
window. The main menu groups the DATA manager and the PLAN and NETWORK focal
points; the Performance, Risk and Cost model sections; and housekeeping core tools such as
the User Manager and Data Type Manager.(2)
Data is managed, imported and exported through the DATA manager, which is in many
ways a starting point for exploring and interacting with the software’s environment. There
is a file system organized in folders and files. Files may be added (uploaded), while new files
are created through adding a pre-defined data table. The software is used by adding and/or
managing data files through the DATA manager, and then using those files in the various
analysis modules available.
Application modules are implemented as individual plug-ins, taking advantage of the platform’s infrastructure (DATA manager, User Manager, Data Type Manager). The anchor
modules, natively present in AWARE-P, are PLAN and the core NETWORK tool. The latter
is the main vehicle for the software to interact with, and express results relative to, the water
network. The target of the AWARE-P analysis is the PLAN tool.
The metrics used in PLAN are grouped in the Performance, Risk and Cost sections of the
main menu. Each of those tools may be used independently. Please refer to the appropriate
sections for further explanation.
Further reading
Alegre, H., Almeida, M.C., Covas, D.I.C., Cardoso, M.A., Coelho, S.T. (2011). Integrated
approach for infrastructure asset management of urban water systems. Proc. IWA LESAM
2011, Germany.
Alegre, H., Covas, D. (2010). Water supply infrastructure asset management – a rehabilitation-based approach. (in Portuguese). Technical Guide no.16. ERSAR, LNEC, IST, Lisboa,
472 pp. (ISBN: 978-989-8360-04-5).
1. In order to make full use of the 3D visualization capabilities on offer, the Google Earth® browser
plug-in should be installed. Additionally, WebGL-capable browsers, such as Google Chrome®, Mozilla
Firefox® or Apple Safari®, should have that feature enabled.
2. Available to system administrators.
p. 12 / 38
Almeida, M. C., Leitão, J. P., Martins, A. (2011). Incorporating risk in infrastructure asset
management of urban water systems. Urban Water (submitted).
AWARE-P (2011). www.aware-p.org
Marques, M. J., Saramago, A. P., Silva, M. H., Paiva, C., Coelho, S., Pina, A., Oliveira, S.
C., Teixeira, J. P., Camacho, P. A., Leitão, J. P., Coelho, S. T. (2011). Rehabilitation in Oeiras
& Amadora: a practical approach. Proc. IWA LESAM 2011, Germany.
See also
• PLAN – AWARE-P Planning
• NETWORK-EPANET
• PI – Performance Indicators
• PX – Performance Indices
• FAIL – Failure Analysis
• CIMP – Component Importance
• UNMET – Expected Unmet Demand
• IVI – Infrastructure Value Index
• CORE
p. 13 / 38
p. 14 / 38
TOOLS
PLAN – AWARE-P planning
Purpose
PLAN embodies the central planning
framework of the AWARE-P infrastructure
asset management methodology, where
planning alternatives or competing projects
are measured up and compared, through
selected performance, risk and cost metrics,
using interactive numerical and 2D/3D
graphical information display.
Overview
PLAN provides an organized assessment
and comparison environment where a
number of competing projects or alternative
designs can be pitched against each
other and numerically as well as visually
compared, with a view to supporting
decision-making.
The tool is based on the 3 main axes that
characterize the assessment and comparison
exercise: a number of alternatives or
projects, a set of standardized metrics and
a given time frame. The latter comprises
a number of user-specified time steps and
includes both a planning horizon (i.e., the
time frame of the intervention itself) and an
analysis horizon (a longer time frame where
the impact of the intervention is assessed).
The metrics selected by the user, which
may come from the performance, risk
p. 15 / 38
and cost assessment tools present in the
AWARE-P environment, or from external
evaluations as selected by the user, are
standardized as numerical indices and then
categorized as color-coded levels, with the
emphasis on coherent definition by the
user of the target category values. A score is
also calculated as a weighted average of the
indices for each alternative, and a ranking
of the alternatives is given according to this
score.
PLAN uses a very flexible 2D/3D cube
display to give the user total control of
which dimensions and viewpoints are
required for analysis.
Usage
PLAN is launched from the AWARE-P
main menu. The initial screen displays any
existing plan files, and gives the option to
create a new plan. Existing plans may be
edited by clicking on the file name, skipping
the creation stage and leading directly to the
tool’s main window.
Creating a new plan entails identifying
the plan name, start year, planning horizon
and analysis horizon. There is optional
space to register planning objectives and
any relevant notes. Completing the required
fields and pressing Create takes the user to
the tool’s main window. A base alternative
(named “Status quo” by default, though this
can be edited) is automatically created.
PLAN’s main window has 3 tabs: Data,
Ranking and 3D Cube. The Data tab is
essentially used to define alternatives and
metrics, and then fill out the resulting
table with values of those metrics for each
alternative, for each of the time steps
included in the analysis. The Ranking tab
is used for comparing and ranking the
alternatives, using essentially 2D views. The
3D Cube tab is used for a tridimensional
display of the results (alternatives, metrics
and time steps).
In the Data tab, time steps can be edited
and further adjusted by pressing the Edit
Plan button. Add Alternative gives access to
the alternative editor, where a code, a name
and a description are filled out. Add Metric
allows the user to specify the type of metric
(performance, risk or cost) as well as a code,
a name(3) and its description.
The most crucial feature in the metric
specification is the set of values that
transform it into a standardized index from
0 to 3, and into color-coded levels, whose
values have the following meaning:
• (Green) 2 – 3: good
• (Yellow) 1 – 2: fair
• (Red) 0 – 1: poor
The user must define the limits of the
green, yellow and red bands for the metric.
A specific weight can be assigned to the
new metric using 5 levels, from very low
to very high (numerically, from 0.5 to 2.0).
The weight can also be specified for each
time step, if the importance of the metric is
thought to vary with time.
The metric may also be marked as
mandatory: if the metric is in the red for
a particular alternative , then the whole
alternative will be ranked in the red,
regardless of how it fares in the other
metrics.
The planning table that is displayed in the
data tab reflects the standardization of the
metrics’ values into indices, by displaying
the standardized values (0-3) in shaded
typeface under each metric value. The
colors reflect the level.
Important: remember to save the
planning table each time it is edited.
The Ranking tab gives access to a colorcoded display of results in 2D tables, for a
specific year (alternatives vs. metrics), for
a specific alternative (metrics vs. years) or
for a specific metric (alternatives vs. years).
Click on any alternative code, metric code
or year to launch the corresponding table.
The Ranking tab also presents the
scores and the respective rankings of
the alternatives, for each year or for each
metric. Overall scores and rankings are
presented when selecting the option Overall
in the Metric drop-down menu. This score
p. 16 / 38
is obtained calculating the weighted mean
of the standardized indices for each metric
in each time step. Metrics that are defined
as mandatory are depicted by a spiky
circle symbol instead of the regular circle.
Different weights associated to the metrics
are translated by the size of the circles.
The 3D Cube tab gives access to a
tridimensional display cube that combines
the 3 views. Full 3D navigation, zooming
and panning is available. Clicking on any
alternative code, on any metric code or on
any year isolates the respective 2D view.
Further reading
Alegre, H., Almeida, M.C., Covas, D.I.C.,
Cardoso, M.A., Coelho, S.T. (2011).
Integrated approach for infrastructure asset
management of urban water systems. Proc.
IWA LESAM 2011, Germany.
Alegre, H., Covas, D. (2010). Water supply
infrastructure asset management – a
rehabilitation-based approach. (in
Portuguese). Technical Guide no.16.
ERSAR, LNEC, IST, Lisboa, 472 pp.
(ISBN: 978-989-8360-04-5).
Marques, M. J., Saramago, A. P., Silva,
M. H., Paiva, C., Coelho, S., Pina, A.,
Oliveira, S. C., Teixeira, J. P., Camacho,
P. A., Leitão, J. P., Coelho, S. T. (2011).
Rehabilitation in Oeiras & Amadora: a
practical approach. Proc. IWA LESAM
2011, Germany.
See also
•
•
•
•
•
•
PI – Performance Indicators
CIMP – Component Importance
IVI – Infrastructure Value Index
3. It is good practice to include the units in the
metric name, encased in brackets, so that
they are displayed in the main planning
table – e.g., “Expected unmet demand (m3/
year)”
NETWORK – EPANET
Purpose
NETWORK–EPANET offers an efficient,
Java-implemented Epanet simulation engine
and natively integrated MSX library, for fullrange hydraulic and water quality network
simulation. It takes advantage of Baseform
Core’s NETWORK and its 2D / 3D network
and results visualization.(4)
Overview
NETWORK makes available a network
simulation engine that is a full Java re-write
and implementation of the Epanet standard.
It integrates the Epanet MSX advanced
water quality simulation library, and offers
the full network modeling functionality,
performing static or extended-period
simulation on .INP standard model files.
The simulator’s implementation is
particularly strong in network visualization,
with the capability to seamlessly overlay
on a range of publicly available maps (such
as Google®, Bing® or OpenStreetMaps) or
on the user’s own maps. It takes advantage
of powerful charting tools that allow for
full manipulation of the network’s values
(parameters or simulation results), and it
uses the Baseform Core’s NETWORK 2D
and 3D network displays to full effect.
Details
Water distribution simulation modeling
details can be found in the extensive
p. 17 / 38
documentation available for the USEPA’s
Epanet simulator (e.g., Rossman, 2000). The
network model implemented is a rigorous,
from-scratch Java implementation of the
Epanet hydraulic and water quality model
and of the Epanet MSX advanced water
quality library (EPA, 2008). A degree of
visual and functional similarity with the
original was sought in integrating it in the
Baseform and NETWORK user interface,
while taking advantage of the latter’s nextgeneration capabilities, such as the 3D
visualization.
At the present stage, network editing is
not available and the program is essentially
intended for use with network model files
prepared elsewhere. Some simulation options
such as the time options or the unit system
can be parameterized, as detailed in Usage.
Usage
NETWORK opens to display a list of
the .INP network model files present
in the selected folder. Clicking on a file
opens it in the main NETWORK window.
Alternatively, from the DATA manager,
clicking on a .INP file name will open a
preview and offer the possibility to open the
model file directly in NETWORK.
Three sets of functionality are available
in separate tabs: Model, Chart & Scale,
and Visualization. On the left-hand side, a
number of collapsible drawers are used in
each tab for setting specific options.
The Model tab displays the content and
settings of the network model contained
in the .INP file opened. It offers a network
summary; a choice of base map layers, such
as Google® or OpenStreetMaps; simulation
time parameterization; easy inspection of
element properties and simulation settings;
and export to .INP, Excel®, XML/KML
formats.
The Chart & Scale tab is where the scales
used in displaying pipe and junction data
and results in the Visualization tab are
selected. Additionally, it uses a (transposed)
cumulative distribution chart that is very
useful in learning about the population of
values present for the particular parameter
or analysis result. The latter are simulationtime sensitive, and therefore a timeline
slider is made available.
The Visualization tab is where 2D/3D
display of the network, its parameters and
modeling or analysis results takes place. 2D
is implemented in a familiar Epanet-like
format; 3D is offered both in a fully embedded Google Earth® visualization, taking
advantage of that platform’s wealth of maps
and 3D features (such as buildings), and in
Baseform’s own high-performance WebGL
3D visualizer. This tab has a full screen
mode for better spatial display capabilities.
The Model tab
The Model tab displays the content and
settings of the network model contained in
the .INP file opened. The Summary gives a
first digest of the model’s main figures, and
includes a button for opening the file in the
DATA manager (this is a recurring feature
in any application tool).
The Layers section allows for selection of
which network features to display, and what
background layer to use for display - sources
available include several Google® and Bing®
layers, as well as OpenStreetMaps®. In case
one of these is selected, the EPSG projection
code(5) for the area concerned must be
specified in order to adjust for the source
file projection on the map.
The Simulation section allows for the
time-related settings of the simulation
to be adjusted, and the simulation to be
launched. Element Properties gives access
to the properties specified for any network
component. Likewise, the Settings section
displays the general simulation settings.
Though the current version does not allow
for editing of these properties, the feature
is expected to be included in forthcoming
versions.
However, the units system and the flow
units can be specified and changed on the
fly, for any network model. This useful
feature does not compromise the integrity
p. 18 / 38
of the .INP model file, as all necessary
changes are taken care of internally by the
implementation: the model file does not
need to be modified for the model to work
in the new choice of units.
Finally, the Export section allows for the
export of model files to INP, Excel® and
XML formats. There is also an export to
Excel® of the full simulation results, as well
as of the visualization data to Google Earth®
/ KML. The Excel option is particularly
useful as a further editing facility and model
section workflow manipulation.
The Chart & Scale tab
The Chart & Scale tab is where the pipe
and junction data to be displayed in the
visualizer are selected, and the respective
scales are adjusted. This includes network
features and simulation model results, as
well as results from any of the network-level
tools present in the AWARE-P portfolio,
such as CIMP or PX.
A (“transposed”) cumulative distribution
chart is used for displaying the population
of values present for the particular parameter or (simulation time-sensitive) modeling
result. Two preset scale modes are available,
by dividing the population into 25% quartiles
in either the X or Y axes. Alternatively, the
scale markers on the chart can be displaced
individually by the user for fine adjustments.
When adjusting the display scales for
time-sensitive results (any network simulation results in an extended-period simulation), care should be taken in selecting the
appropriate time frame. For example, if
investigating low velocities, it will be a good
idea to select a time step when demand is
low, such as during nighttime.
The Visualization tab
This is where full 2D/3D visualization
of the network, its parameters and
modeling or analysis results takes place.
The Visualization tab has the capability to
display network features and simulation
model results, as well as results from any
of the network-level analysis tools present
in the AWARE-P portfolio, such as CIMP
or PX. The pipe- or node-related feature to
display is selected in the Chart & Scale tab.
A full screen mode is accessible from this
tab. A time-slider controls any time-sensitive
result display, in a standard way as usually
found in simulators. There is the option to
play the simulation at a video-like experience, in single, double or maximum-available speed. The parameters or results shown
are those selected in The Chart & Scale tab,
with the corresponding selected scale.
Selecting 2D displays the network in an
Epanet-like map. The pipes and junctions
are color-coded to reflect the scale chosen.
If displaying time-sensitive values, such
as model results in an extended-period
simulation, the time slider should be
adjusted to the desired time-step.
3D WebGL displays the network and the
selected parameters in a very fast native 3D
viewer implemented using WebGL browserside technology. A browser that enables
WebGL must be used(6) in order for this
option to be available. In this viewer, it is
possible to use the customary pan and zoom
commands, as well as snap onto predefined
perspective views that include isometric and
North, West, South and East.
3D Google Earth® displays the network
and the selected parameters over Google
Earth® visualization. Usage is intuitive and
will be very familiar to users of Google’s
viewer. It is possible to toggle on/off the
representation of model junctions as well
as place & road names, 3D buildings and
the zoom & scale controls. Turning on
3D buildings provides the most complete
viewing experience, in locations where
building shapes are available in the Google
Earth database. Note that building shapes
may need a few more seconds to load than
the underlying terrain view, depending on
internet connection speed, but are usually
worth the wait in terms of the richness of
presentation of analysis results. This viewing
option requires the Google Earth® plug-in to
be installed in the browser.(7)
p. 19 / 38
Further reading
EPA (2008). EPANET-MSX (Multi-Species
eXtension). http://www.epa.gov/nrmrl/
wswrd/dw/epanet.html#extension
(accessed 2012/01/12)
Rossman, L. (2000). Epanet 2.0 User’s
Manual. Water Supply and Water
Resources Division, National Risk
Management Research Laboratory, U.S.
Environmental Protection Agency.
See also
•
•
•
•
Baseform CORE
4. This model for pressurized networks
is the first in a series that will in the
future hopefully include GIS network file
compatibility and models for wastewater/
stormwater networks.
5. see www.epsg.org and www.epsg-registry.org
6. At the date of publication, browsers offering
that feature include Firefox®, Chrome®,
Opera® and Safari®.
7. Go to http://www.google.com/earth/
explore/products/plugin.html
PI – Performance Indicators
Purpose
PI allows for the quantitative assessment
of the efficiency or effectiveness of a system
through the calculation of performance indicators based on state-of-the-art, standardized PI libraries as well as user-developed or
customized sets.
Overview
Underpinning the AWARE-P methodology and embedding the principles of the ISO
standards 24510/24511/24512, the PI tool
makes available a performance indicators
framework for rigorous assessment of urban
water system efficiency and effectiveness.
The tool allows the users to select PIs
from a rather comprehensive list organized
by objectives and assessment criteria.
The main leading-edge reference libraries
of performance indicators relevant for
infrastructure asset management of urban
water services are incorporated, including
the International Water Association (IWA
PI systems) libraries, the CARE-W and
CARE-S libraries and the Portuguese Water
Services Regulator (ERSAR) libraries.
Other indicators developed within the
AWARE-P project are also included. The
user is free to further edit the database (fully
MS Excel compliant) and customize the
list of objectives, criteria and performance
indicators offered.
The user creates a set of PIs, through
p. 20 / 38
a shopping list mechanism – intuitive
search allows the user to filter out a given
subset of available pre-defined PIs – e.g.,
drinking water or wastewater, IWA/CARE/
ERSAR libraries, a specific assessment
criterion, a given keyword. Once a given
PI is pre-selected, the user is shown its full
definition (code, name, units, concept,
processing rule, comments and input
variables needed) and is given the option
of including it in the shopping list. When
this process is complete, a table is produced
with all the variables needed, ready for
data input referred to one or several userdefined periods of time. The indicators are
automatically calculated.
Details
Performance assessment refers to the
evaluation of the efficiency or the effectiveness of a process or activity through
the production of performance measures.
Performance measures are the specific parameters that are used to inform the assessment (Matos et al., 2003, Alegre et al. 2006,
Cabrera & Pardo, 2008, ISO 24510, ISO
24511, ISO 24512). Performance indicators
are quantitative efficiency or effectiveness
measures for the activity of a utility.
A performance indicator consists of a
value (resulting from the evaluation of a
given processing rule) expressed in specific
units, and a confidence grade which indicates the quality of the data represented by
the indicator. Performance Indicators are
typically expressed as ratios between variables; these may be commensurate (e.g. %) or
non-commensurate (e.g. $/m3). In the latter case, the denominator should represent
one dimension of the system (e.g. number
of service connections; total mains length;
annual costs), to allow for comparisons.(8)
The components of PI systems should
comply with some key requirements (ISO
24500 standards). Performance indicators
are computed from variables, and interpreted
taking into account explanatory factors.
An explanatory factor is any element of the
system of performance indicators that can
be used to explain PI values, at the analysis
stage. This includes PI, variables, context information and other data elements not playing an active role before the analysis stage.
AWARE-P’s PI tool is based on the
concept of PI libraries, coherent sets of
PI developed for a specific purpose, from
regulation, international statistics, global
management of the utility or for a given
decision support system. The PI tool
includes libraries from some the world’s
most relevant PI systems in the field of
urban water supply and wastewater/
stormwater services, such as those
developed by the International Water
Association (IWA PI systems), by the
CARE-W and CARE-S projects and by
the Portuguese Water Services Regulator
(ERSAR), one of the most advanced
regulatory systems internationally.
Besides those sets of highly validated,
professionally developed PI, the PI tool allows the user to modify, customize or define
own PI or libraries, and easily share or interchange them in the shape of Excel files.
Usage
The PI tool is started from the main
menu. The first step when creating a new
PI analysis file is to select a PI library (e.g.
IWA, AWARE-P), which contains a set of
performance indicators. The PI tool main
window presents all available PIs in the
selected PI library, organized by the objectives and criteria proposed in the AWARE-P
methodology (see page 10).
By selecting the chosen PIs, in accordance
with the objectives and assessment criteria, a
list of the variables needed for their calculation is automatically generated. The following
step is the definition of the timesteps, i.e. the
instants in time, in which the PIs are calculated. The value of each input variable involved
in the calculation of the PIs is introduced by
the user, for each timestep. By pressing the
Save button, the software will calculate the
values of the PIs in each timestep.
p. 21 / 38
The PI libraries available in the software
are to be taken as a reference, but the user
may decide to incorporate new objectives,
assessment criteria or even performance
indicators that may be deemed better
suited to a particular need. To do so, it is
possible to export (download) a predefined
PI library from the Data Manager into a
.xlsx file. This file is easily edited in Excel®,
enabling the user to modify existing
PIs, variables or criteria, and create new
ones – the user must only make sure the
table’s format is preserved so that it can be
imported (uploaded) again. The following
points are important:
• Each row of the Excel table represents either
a PI (pi_type as PI), an input variable (pi_type
as UI), or a criterion (pi_type as OBJ).
• PI rows must present: PI as pi_type; a unique
pi_code; and a pi_rule that uses only existing
input variables (UI).
• The input variables must present: UI as pi_
type; and a unique pi_code.
• The criteria must present: OBJ as pi_type;
a unique pi_code; a pi_description that
indicates the set of associated PIs, identified
by their pi_code; and a pi_group specifying
the associated objective.
The modified xlsx file can then be imported
(uploaded) into a new Performance Indicator
library, which can be created using the function Add table in the Data manager. It will be
automatically saved in the user’s profile.
Further reading
Alegre, H. (2008). Infrastructure asset management of drinking water and wastewater
systems (in Portuguese), TPI 52, LNEC,
Lisbon, ISBN 9789724921341 (385 p.).
Alegre, H., Baptista, J.M., Cabrera Jr.,
E., Cubillo, F., Duarte, P., Hirner,
W., Merkel, W., Parena, R. (2006).
Performance indicators for water supply
services, 2nd edition, Manual of Best
Practice Series, IWA Publishing, London,
ISBN: 1843390515 (305 p.).
Alegre, H.; Cabrera, E.; Merkel, W. (2008).
Current challenges in performance
assessment of water services. Water
Utility Management International, Vol. 3,
N. 3, IWA Publishing (p. 6-7).
Alegre, H., Cabrera, E., Merkel, W.
(2009). Performance assessment
of urban utilities: the case of water
supply, wastewater and solid waste.
Journal of Water Supply: Research and
Technology—AQUA (JWSRTAQUA-D08-00041R1) n.º 58.5/2009 (305-315 p.).
Cabrera, E., Pardo, M.A. (eds.) (2008).
Performance Assessment of Urban
Infrastructure Services: drinking water,
wastewater and solid waste, IWA
Publishing, ISBN: 9781843391913, IWA
Publishing.
ISO 24510:2007 Activities relating to
drinking water and wastewater services Guidelines for the assessment and for the
improvement of the service to users
drinking water and wastewater services
- Guidelines for the management
of wastewater utilities and for the
assessment of wastewater services Guide
drinking water and wastewater services
- Guidelines for the management of
drinking water utilities and for the
assessment of drinking water services.
Matos, R., Cardoso, M.A., Ashley, R,
Duarte, P., Schulz A (2003). Performance
indicators for wastewater services, Manual
of Best Practice Series, IWA Publishing,
ISBN: 9781900222907 (192 p.).
See also
• PI – Performance Indicators
• PX – Performance Indices
8. The use as denominators of variables that
may vary substantially from one year
to another, particularly if not under the
control of the undertaking, should be
avoided (e.g. annual consumption, that may
be affected by weather or other external
reasons), unless the numerator varies in the
p. 22 / 38
same proportion. The information provided
by a performance indicator is the result of
a comparison (with a target value, previous
values of the same indicator, or values of the
same indicator from other undertakings)
(Alegre et al. 2006; ISO 24500).
Purpose
The PX model produces technical performance metrics based on the values of certain
features or state variables of water supply
and waste/stormwater networks. The indices
measure performance concepts related to
level-of-service, network effectiveness and
efficiency, in areas such as hydraulic capacity,
water quality, redundancy or energy behavior.
Overview
The PX tool produces performance indices and levels by evaluating the numerical
results of network simulation models. It uses
a network model file representing the appropriate set of conditions for the analysis of the
desired network. The PX are evaluated at the
component level and then generalized to a
network-wide value.
The PX are selected from the AWARE-P
extensive library of water supply and waste/
stormwater performance indices, which is
continually updated with the latest R&D
advances in the field, and may be edited,
modified and added to by the informed
user. Examples of PX are:
• compliance with a minimum required service
pressure at network nodes, by comparing
with a user-defined reference and a zeroconsumption threshold;
• compliance with a maximum required travel
time at network nodes, by comparing with a
user-defined maximum reference travel time
p. 23 / 38
and a given tolerance above it.
The results are produced both at pipe/node
level, and at the aggregated network level.
Details
Performance assessment in urban water systems may complement performance
indicators (see PI in p.16) with performance
indices and levels (Alegre & Cabrera, 2011).
Performance indices are measures resulting from the combination of disaggregated
performance measures (e.g. weighted average
of performance indicators) or from analysis
tools (e.g. simulation models, statistical tools,
cost efficiency methods). In general, they aim
at aggregating several perspectives into in a
single measure. Compared to performance
indicators, their main advantages are that they
can be more aggregated measures and can
be used to assess future scenarios (e.g. using
simulation results or statistical analyses).
Performance levels are measures of a
qualitative nature, expressed in discrete
categories (e.g. excellent, good, fair, poor).
In general they are adopted when the use
of quantitative measures is not appropriate
(e.g. evaluation of customer satisfaction by
means of surveys) or when synthesizing and
standardizing a range of different metrics,
e.g. as the basis for a decision–making
process (such as can be found in PLAN).
The model used to calculate performance
indices is a relatively straightforward
process that applies a performance curve to
the values of a given network model result,
e.g., flow velocity (Coelho, 1997, Cardoso
et al., 2004; Cardoso et al., 2005). The value
thus obtained contains a performance
judgment, expressed on a standardized 0-3
scale that implies good (2-3), adequate (1-2)
and inadequate (0-1) ranges.
The performance curve is the fundamental evaluation mechanism and is usually designed by a knowledgeable analyst.
It should reflect the users’ sensitivity, and
is often parameterized to that effect. Each
performance index is associated with one
or several ways to calculate a network-wide
value from the component values. This is
termed a generalizing function, and may
take up the form of a weighted average or a
given percentile (including extremes) of the
component values.
Usage
PX opens to display a list of the
performance index tables available in
the selected folder. Creating a new table
requires selecting an uploaded Epanet
model, where the performance index will
be evaluated. A performance index must
be selected from the available indices of the
selected PX library and the name of the new
performance index table must be filled out.
Pressing Create takes the user to the main
PX window.
A brief explanation of the selected
performance index is given at the top of
the main PX window. In order to calculate
the index of each pipe, the input boxes
of the reference values must be filled
out. Pressing Calculate will calculate the
performance index for each pipe/node,
using the reference values, and presented in
the 2-D network model. The PX results can
be visualized in a 3-D network model, using
the NETWORK tool.
New PXs can be introduced by modifying
the PX libraries. In the Data manager it is
possible to download a Performance Index
library into a xlsx file. Opening this file will
allow the user to modify existing PXs and
creating new ones. The modified xlsx file,
with new PXs, can then be uploaded into
a Performance Index library. As in PI, the
user must make sure that the table’s format is preserved so that it can be imported
(uploaded) again. A new Performance Index
library can be created using the function
Add table of the Data manager.
Further reading
Alegre, H., Cabrera, E. (2011). Performance
Indicators. In WaterWiki, updated
2011/08/05 (http://iwawaterwiki.
org/xwiki/bin/view/Articles/
p. 24 / 38
PerformanceIndicators), IWA.
Cardoso, M. A., Coelho, S. T., Matos,
R., Alegre, H. (2004). Performance
assessment of water supply and
wastewater systems. Urban Water Journal
1 (1), pp. 55-67.
Cardoso, M. A., Coelho, S. T., Praça. P.,
Brito, R. S., Matos, J. (2005). Technical
performance assessment of urban sewer
systems. J. Performance of Constructed
Facilities 19 (4), ASCE, pp. 339-346.
Coelho, S.T. (1997). Performance in
water distribution: a system’s approach,
Research Studies Press -John Wiley &
Sons, New York, E.U.A. (225 p.).
Coelho, S.T., Jowitt, P.J. (1997).
Performance analysis in water
distribution, Computing and Control
for the Water Industry, Research Studies
Press - John Wiley & Sons, New York,
E.U.A. (pp.3-20).
See also
• PI - Performance Indicators
• NETWORK-EPANET
Purpose
The aim of the FAIL model is to predict
future pipe or sewer failures for a given
network, e.g. in the context of estimating
risk or cost metrics. It requires an organized
failure history to be supplied, in the form of
work orders and pipe data records, in order
to predict future behavior.
Overview
Two alternative stochastic processes are
offered for calculating failure predictions:
the Poisson process and the Linear Extended
Yule Process (LEYP). The probability distribution of the number of failures is estimated
using the maximum likelihood method.
The probability of failure and the number of
future failures is predicted for each pipe, using the probability function and the expected
value of the stochastic process selected.
The failure data must be provided in two
data tables: (i) a work orders (maintenance
records) table; and (ii) a pipe inventory table,
containing the universe of pipes that the work
orders table refers to. This is not necessarily
the same universe of pipes – i.e., network or
sector – that the results of the analysis will be
applied to: often the latter is a subset of the
former; it may be a distinct set altogether, e.g.,
if the results from one system are applied to
another where data is not available.
In the pipe inventory table, each pipe
is identified by a unique IPID code and
p. 25 / 38
described by length, installation year,
material and diameter. The maintenance
records table gathers all failure records
of the water network. Each record must
be associated with a pipe IPID code and
contain failure date. These two tables are
associated in order to build the failure
history of each pipe.
The analysis is applied to a network that
must be specified in the form of a network
model file (.INP format). A summary of
network information is displayed, along
with a network map.
The NETWORK tool and its visualization
capabilities may be invoked from this tool,
namely as a swift shortcut to visualize the
results on 2D or 3D maps – results become
available for display in that tool as soon as
they are produced.
Details
The Poisson process
A Poisson process is a counting process
in which the events occur independently
at a constant rate and where the number
of events follows a Poisson distribution.
It is assumed that the rate of the counting
process is proportional to the length of each
pipe. The failure rate is estimated by the
maximum likelihood method.
If the failure rate were estimated using the
entire data set, then it would be the same
for all pipes, no matter their properties. In
this implementation, the data set is divided
based on the pipe material, thus creating
various pipe material categories. Once the
pipe data set is categorized, the failure rate
is estimated for each category using the
maximum likelihood method.
The predicted number of failures in each
pipe is obtained using the expected value of
the Poisson distribution, whereas the failure
probabilities are obtained using the Poisson
probability function.
The Linear Extended Yule Process (LEYP)
The Linear Extended Yule Process (LEYP)
implemented in this project is a counting
process where the intensity function
depends on the age of the pipe, the number
of past events and a vector of covariates
(potentially predictive variables, such as
pipe diameter), (Le Gat, 2009; Martins,
2011). The covariates taken into account in
this implementation are the pipe diameter
and the logarithm of pipe length.
Furthermore, the data set is divided
according to the material of each pipe. For
each pipe material category, there will be
a different set of estimated parameters.
All parameters are estimated through the
numeric maximization of the log-likelihood
function, derived by Le Gat (2009), using
the Nelder-Mead nonlinear optimization
method. A significance test is carried out for
each estimated parameter, resorting to the
likelihood ratio test, using the Chi-square
distribution, given by the Wilks theorem
approximation.
Once the LEYP parameters have been
estimated, the failure probabilities of
each pipe are obtained using the Negative
Binomial probability function, presented
in Le Gat (2009) and Martins (2011). The
predicted number of failures is obtained
as the expected value of the same Negative
Binomial distribution.
For further details about the theoretical
models behind both processes, and their
implementation in the software, please refer
to Appendix A.
Usage
FAIL is launched from the Failure Analysis option on the AWARE-P main menu.
The initial screen displays any existing
failure analysis tables, and gives the option
to create a new table.
Creating a new failure analysis table
entails selecting an existing work order
pipes table and an existing work order
failures table, which can be uploaded in the
DATA tool. Furthermore an Epanet model
can be selected in order to visualize the
estimated failure probabilities.
The main FAIL window presents a pipe
inventory description; failure estimates for
p. 26 / 38
each pipe material; failure estimates for
each pipe; and a visualization of the failure
probabilities in the uploaded Epanet model
network.
In the middle of the FAIL window an
option button allows to switch between two
stochastic models to estimate the failure
probabilities: Poisson or LEYP.
When the LEYP option is selected, under
each estimated parameter, the associated
p-value is presented. If the p-value is close
to zero, the associated parameter is more
significant.
Selecting a network model on the
left-hand side menu (or if a model was
previously selected) will allow to visualize
the failure estimates on a 2-D map.
Importing failure data
The failure analysis tool requires a work
order pipes table and a work order failures
table. Both tables can be created in the
Data manager tool, selecting the Add table
option. In the Add table window the name
of the table must be filled out and a table
type must be selected (work order pipes or
work order failures). Pressing the button
Create will add a new empty table, with 0
rows, to the specified folder.
In order to import the required failure
data, both empty tables can be downloaded
as xlsx files, filled out with the necessary
information and uploaded next.
The six pipe attributes to be filled out in
the work order pipes table are:
• pipe_id: an alphanumeric code that identifies
uniquely each pipe of the water network.
Does not accept empty values;
• material: a text value representing the pipe
material;
• diameter: a numeric value representing the
pipe diameter. Does not accept empty values;
• length: a numeric value representing the pipe
length in meters. Does not accept empty
values;
• installation_date: a date value representing
the installation date of each pipe, in the form
day-month-year (e.g. 27-03-1994). Does not
accept empty values;
• decommissioning_date: a date value
representing the decommissioning date of
each pipe, in the form day-month-year (e.g.
27-03-1994).
The four failure attributes to be filled out
in the work order failures table are:
• failure_date: a date value representing the
date of occurrence of each failure, in the form
• failure_type: a text value representing the
type of failure (e.g. break or leakage);
• failure_duration: a numeric value representing
the downtime caused by each failure;
the pipe where each failure occurred; does
not accept empty values.
Further reading
Le Gat, Y. (2009). Une extension du
processus de Yule pour la modélization
stochastique des événements récurrents.
Application aux défaillances de
canalizations d’eau sous pression. Ph.D.
thesis, Cemagref Bordeaux, Paristech.
Martins, A. (2011). Stochastic models for
prediction of pipe failures in water supply
systems. MSc thesis, Instituto Superior
Técnico, Technical Univ. Lisbon, Portugal
Martins, A., Amado, C., Leitão, J.P.
(2011). Stochastic models for prediction
of pipe failures in water supply systems.
(undergoing submission)
See also
•
•
•
•
NETWORK-EPANET
PLAN – AWARE-P Planning
p. 27 / 38
CIMP – Component importance
Purpose
The CIMP model calculates a component
importance metric for each individual pipe
in a network, based on the impact of its
failure on nodal consumption. The measure
is computed based on the network’s hydraulic model, using full simulation capabilities.
Component importance (also termed hydraulic ‘criticality’) is a crucial measure of a
pipe’s consequence in the network, used for
example in the assessment of risk associated
with pipe failure.
Overview
The component importance of each individual pipe is calculated by comparing the total
demand that the network is hydraulically able
to satisfy when that pipe is out of service, with
the total demand that the original network is
able to supply. The calculation is computed
over the entire simulation duration specified
in the network model used – i.e., the unmet
demand caused by each individual pipe failure
is added for all time steps and compared with
the total supplied by the original network over
the entire simulation duration.
Component importance values are given,
for each pipe, between the values of zero
(i.e., if the pipe fails, all network demand is
still satisfied, over the simulation duration)
and 1 (i.e., when the pipe fails, no demand is
satisfied across the entire network, over the
simulation duration).
For example: if a given pipe has a
component importance of 0.81, it means
that, when the pipe fails, the network
will not be able to supply 81% of the total
demand (i.e., only 19% will be supplied).
In addition, the actual value of the unmet
demand over the simulation period is
shown for each pipe.
they are produced.
Details
The calculation of satisfied demand
(actual consumption) is based on a simple
relationship between available pressure and
effective consumption for the particular
simulation time step at each node. This
relationship is built on two user-specified
reference pressure values:
• the Zero-Consumption Pressure is the
value below which there is no physical
consumption at the node (e.g., 8 m / 24 ft);
and
• the Required Minimum Pressure is the
nodal pressure value above which the nodal
demand is considered to be fully satisfied
(e.g., 25 m / 75 ft).
A linear interpolation is used for pressure
values in between the two limits. Nodal
demand is understood as the specified basedemand multiplied by the demand pattern’s
factor and by any applicable demand
multiplier.
The computation is based on full hydraulic response simulation as provided by the
network model, where the nodal pressure
values for each time step are computed for
the reduced network (i.e., with the target
pipe missing), and the expected satisfied demand at each node is calculated by applying
the above relationship. The total demand for
the network, which is used as the basis for
the ratio, is computed in the same way but
with the original network. The current ver-
p. 28 / 38
sion uses Epanet’s standard demand-driven
hydraulic model.
Usage
CIMP is launched from the AWARE-P
existing component importance tables,
and gives the option to create a new table.
Creating a new component importance
table requires an uploaded Epanet model.
A brief explanation of the tool is given at
the top of the main CIMP window. In order
to calculate the component importance of
each pipe, the two input boxes Zero-Consumption Pressure and Required Minimum
Pressure must be filled out (default values
are 15.0 m and 35.0 m, respectively). Pressing Calculate will compute the percentage
of unmet demand caused by the closing of
each pipe of the network.
Further reading
Andrianov, A. (2010). MIKE NET and
RELNet: which approach to reliability
analysis is better? Available at: http://
www.vateknik.lth.se/exjobb/E315.pdf
[accessed: 19 July 2010]
CARE-W., 2003. Tests and validation of
Technical Tools. Cemagref, INSA Lyon,
NTNU, Brno University. Report.
CARE-W., 2004. Guidelines for the use
of Technical Tools. Cemagref, SINTEF,
INSA Lyon. Report.
Wagner, J. M., Shamir, U., Marks, D. H.
(1998). Water Distribution Reliability:
Simulation Methods. Journal of Water
Resources Planning and Management,
114(3), pp. 276-294.
See also
•
•
•
•
NETWORK-EPANET
UNMET –
Expected Unmet Demand
expected loss of service or loss of revenue (if
multiplied by the unit revenue).
Purpose
The UNMET model calculates a service
interruption risk metric expressed as the
expected volume of unmet demand in a
system over one year, given the expected
number of outages for each pipe, the
average downtime per pipe outage, and
the component importance of each pipe,
expressed in terms of unmet demand.
Usage
Overview
The tool combines the results of the Failure
Analysis and the Component Importance
tools (although it can use failure rate and
component importance tables manually produced in the same format).
they are produced.
Details
For each pipe, the value of expected unmet
demand in case of outage is multiplied by the
pipe’s expected number of failures in 1 year,
and by the average outage time (user input).
The result is the expected value of the
total volume of unmet demand in the
network caused by the individual outage of
each pipe. This provides a direct measure of
p. 29 / 38
See also
•
•
•
•
NETWORK-EPANET
CIMP – Component Importance IVI – Value Index
Purpose
The Infrastructure Value Index represents
the ageing degree of an infrastructure, and
is calculated through the ratio between the
current value and the replacement value of
the infrastructure.
they are produced.
Details
The IVI at a given time t (IVIt) is a performance-cost measure that reflects the age
of an infrastructure. It is given by the ratio
between the present value and the respective
replacement value (Alegre & Covas, 2010) of
the infrastructure. This index is particularly
adequate for establishing goals associated to
infrastructural sustainability criteria.
The calculation of the IVI starts with the
calculation of the residual life for all pipes
and associated present value, individually.
Subsequently, the global infrastructure
IVI is calculated (Equation 8). The IVI of
a pipe corresponds to the percentage of its
remaining life.
(8)
Overview
This tool calculates the Infrastructure
Value Index. The process of calculation of IVI
is based on individual pipe characteristics,
but the IVI is presented to the infrastructure
as a whole.
The cost data used to calculate the IVI
must be provided in one data table: a cost
table. The cost table contains the characteristics of the infrastructure pipes that are used
to calculate the global infrastructure IVI.
In the cost table, each pipe is identified
by a unique IPID code and described by
length, installation year, material, diameter,
useful life time and construction and
replacement cost.
For the specific case of a water supply
system infrastructure, the analysis can
also be applied to a network that must be
specified in the form of a network model
file (.INP format). A summary of network
information is displayed, along with a
network map.
p. 30 / 38
Where:
• IVI(t) is the dimensionless infrastructure value
index at year t (-);
• t is the evaluation year (year);
• N is the number of infrastructure components (-);
• csi,t is the replacement cost of the
infrastructure component i in year t (currency
units);
• vri,t is the residual lifetime of infrastructure
component i in year t (year); and
• vui is the total technical lifetime of
infrastructure component i (year).
Infrastructure value index values for
stabilized infrastructures should be around
0.5 (e.g. 0.4-0.6), i.e. the investment
during a specific time period is equivalent,
in average, to the depreciation of the
infrastructure during the same time period.
IVI values above 0.6 can mean one of the
following:
• recent infrastructures not yet stabilized;
• new developments in old infrastructures; or
• infrastructures where there has been a
rehabilitation over-investment.
Low IVI values (e.g. <0.4) can mean
a deteriorated infrastructure, requiring
significant rehabilitation investment.
Usage
IVI is launched from the AWARE-P
existing cost tables.
New cost tables are created in the
DATA tool, in which pipe cost data can be
uploaded.
The main IVI window presents a
summary of the cost table and information
about the Epanet network model that can
be associated to the cost table on the lefthand side menu, and a summary of the pipe
inventory per pipe material: number of
pipes, total length, total construction cost,
total replacement value, the ratio between
the residual life and useful life, present
value and IVI; IVI for each pipe; and a
visualization of the IVI in the associated
Epanet network model.
Selecting a network model in the
left-hand side menu (or if a model was
previously selected) allows visualising the
IVI on a 2-D map.
Importing cost data
The IVI tool requires a cost table; this
table is created in the DATA manager tool,
selecting the Add table option. In the Add
table window the name of the table must
be filled out and the cost table type must
be selected. Pressing the button Create will
add a new empty table, with no rows, to the
specified folder.
In order to import the required cost data,
the empty cost table can be downloaded
as xlsx file, filled out with the necessary
information and then uploaded (if a filled
in cost table is already available, it can be
directly uploaded without downloading the
empty cost table).
The eight pipe attributes required for the
cost table are:
uniquely each pipe. Does not accept empty
values;
p. 31 / 38
• pipe_material: a text value representing the
pipe material;
• pipe_diameter: a numeric value representing
pipe diameter. Does not accept empty values;
• pipe_length: a numeric value representing
the pipe length in meters. Does not accept
empty values;
• installation_date: a date value representing
the installation date of each pipe, in the form
• pipe_usefullife: an expected (design) lifetime
of each pipe. Does not accept empty values;
• construction cost: per unit length of pipe.
Does not accept empty values;
• replacement cost: per unit length of pipe.
Does not accept empty values.
Further reading
Alegre, H., Covas, D. (2010). Infrastructure
Asset Management of Water Services.
An aproach based on rehabilitation (in
Portuguese). Technical Guide 16. ERSAR,
LNEC, IST, Lisboa, 472 pp. (ISBN: 978989-8360-04-5
See also
• PI - Performance Indicators
CORE
Purpose
The nuclear functionality of a platform
that has been built from scratch with pluggability and extensibility in mind, Core is the
conceptual and technical conductor of every
app and tool in the baseform portfolio.
Overview
Core is a common set of functionality and
services used by baseform tools, including
data manager, data type manager, webenabled user interface, user manager and
network/data visualization environment.
Every baseform tool exists on top of
baseform core, using these features and
working within defined boundaries.
Core is the reason why going from one
tool to the next is a seamless experience, all
inside the same platform, always knowing
where things are and how they behave.
p. 32 / 38
Details
Tools & apps
Baseform is a host of different functionality meant to work together. Baseform tools
are plugins to the core platform. Baseform
applications (apps) are suites of tools working together to provide added functionality.
Using an example from a well-known
application family: Microsoft Word®
includes a charting tool which is also
present on Microsoft Excel®; this tool makes
sense on both applications, as it provides a
clear function and works in a common way
on both.
Data manager
Data manager works like Windows
Explorer® (or the Finder® if you prefer the
Mac): you have files and folders there, you
can organize, sort, rename, copy, move, etc.;
all in a familiar way.
Data manager is actually smarter than
a regular file manager. Data is defined as
having two possible forms: data files and
data tables. Data files are what you would
expect: images, pdfs, binary outputs,
etc.. Data tables are seen as special data
that is organized in rows and columns as
any spreadsheet; in fact, baseform core
recognizes spreadsheets as particularly
important tools, and allows for the import
and export of any data table to and from
native MS Excel® files.
It is smarter because it knows specific
data types, and acts on them in specific
ways. For example, it knows a data table
of the ‘failure analysis’ type (containing
the probability of failure of each pipe of
a system); it lets you look at the data in a
tabular way, but it also invokes the network
visualizer to show an image of the network
with the data.
Each data type can have a specific
manager allowing you to see, create and
interact with data in a special, direct and
meaningful way. You can even use our data
type manager to extend current data types
to your specific needs, or to create new data
types altogether.
Finally, data manager recognizes that data
is interdependent. It registers that a certain
file was used to create another; it shows you
that and lets you navigate through interdependent files and tables. One way this
makes a lot of sense is that you can backup
one data file, including its all dependent
data, analysis and tables from one instance
to another just by clicking one button,
preserving all the dependencies.
Users: accountability, security and
permissions
The baseform platform is designed to
be just as good for one user as for multiple
users, and to gather as many users from any
organization or project as possible inside
every application. Thus, it is natural to provide user, security and permissions management. Baseform core has built-in user management that is both simple and powerful.
The data manager allows users to manage
their files and share permissions, and every
tool has access to this security framework.
Security also means accountability: by
default every action is duly logged; access
and searching in these logs is given inside
the applications to super/admin users.
As many other aspects of baseform core,
user management is there but you use it
only if you need it. If all you need is just to
download an app to run on your computer,
you will not be bothered by it.
Networks: infrastructures, systems and
data from a new angle
Baseform is a platform designed for hosting a growing family of tools and applications for networked urban infrastructures.
This system approach means that accessing
and relating to network files is central for
every tool.
Cutting-edge research and development
is applied to interacting with data and especially with network data. You can chart
any parameter, variable or assessment; you
can see your networks over any maps, or on
top of interactive Google Earth®, and click
on any element; you can explore our unique
3D viewing and you can even play back and
p. 33 / 38
forth with time to visualize network behavior.
Whenever a new tool is developed, making
available new assessments for your system,
it will directly benefit from baseform’s core
interactive network visualization.
One way the baseform platform is moving
forward is by expanding the possible
definitions of ‘network’: currently, it has
world-class support for hydraulic models
using Epanet, but incorporation of GIS and
other network models is on its way..
Inclusive technology: running in the
present and in the future
Core defines a new technology environment designed for running everywhere.
Baseform apps run on (just a few examples):
• MacBook Air
• Any Windows PC, including netbooks such
as the Eee line
• Any virtual machine, Blade or Unix server
environment
• Regular or private cloud servers, such as the
Amazon EC2 platform.
Those were examples of configurations
able to run baseformed apps. In order to use
them, all you need is a modern web browser:
• Full functionality, including 3D WebGL, is
available on Chrome, Firefox or Safari, on
Windows, Linux or Mac OS.
• Main functionality also works fine in Android
smartphones and tablets, and on iPhones
and iPads.
Appendix A
Notes on the theoretical models used in FAIL The following notes provide a brief introduction to the theoretical models behind the techniques used in the FAIL tool Notation The number of failures through a time period t constitutes a counting process {! ! , ! ∈ ℝ! }. !(!) is the number of failures in the time interval [0, !]; ! ! − ! ! is the number of failures during [!, !]; ![!(!) − !(!)] is the expected number of failures during [!, !]; !{!(!) − !(!) = !} is the probability to fail � times during [!, !]; ! ! . is the likelihood function of the process. The Poisson process A Poisson process is a counting process in which the events occur independently with a constant rate ! and where the number of events follows a Poisson distribution, i.e. ! ! ~!"#$$"%(!"). It is assumed here that the rate of the counting process is proportional to the length of each pipe. The number of failures in pipe ! follows the counting process. {!! ! , ! ∈ ℝ! }, with rate !! = !!� , where !! represents the length of the pipe and ! represents the failure rate per km (!"#$%& !" !"#$%&'(/ !"#$/ !") (Martins et al., 2011). ! is estimated by the maximum likelihood method and presented in Equation (1). !
!!! !!
!
!!! !! !!
λ = argmax!∈ℝ! L λ n, t, l =
where: , (1) ! is the number of pipes; ! = [!! … !! ] and !! is the number of recorded failures of pipe ! ; ! = [!! … !! ] and !! is the length of pipe ! ; ! = [!! … �! ] and !! is the observation period of pipe ! . If ! is estimated using the entire data set, then the estimated failure rate will be the same for all pipes, no matter their properties. Nevertheless, the data set can be divided based on the pipe characteristics, such as material and diameter, creating different categories. Once the pipe data set is categorized, the failure rate !! can be estimated for each category !! using Equation (1), restricted to the pipes in !! . The probability of a pipe ! in category !! to fail ! times during the time interval [!, ! + !] is given by Equation (2). P N t+s −N t =n =
! ! ! ! ! ! ! ! !! !
!!
!
The predicted number of failures in pipe ! is based on the expected value of the Poisson, Equation (3). E N t + s − N(t) = λ! l! s AWARE-P software documentation: draft version 2012-02-02
p. 34 / 38
(2) (3) The Linear Extended Yule Process (LEYP) The Linear Extended Yule Process (LEYP) implemented in this project is a counting process where the intensity function ! (!, !, !) depends on the age of the process, the number of past events and a vector of covariates (potentially predictive variables, such as pipe diameter), as in Equation (4) (Le Gat, 2009; Martins, 2011). ⊺
γ t, j, x = 1 + αj δt !!! e! ! (4) where: ! = 1 !! … !! represents the vector of ! covariates; ! = !! !! … !! is the vector of parameters associated with the covariates; ! is the number of previous failures; ! is the parameter associated with the number of previous failures (when ! = 0 the rate is independent of the number of previous failures); ! is the parameter associated with the age of the process (when ! = 1 the rate is independent of the age of the pipe). The covariates taken into account in this implementation are the pipe diameter and the logarithm of pipe length. Therefore, ! = 1 !!"#$ !!" !"#$%! and ! = !! !!"#$ !!" !"#$%! . Furthermore, the data set is divided according to the material of each pipe. For each pipe material category there will be a different set of estimated parameters ! , ! and !. All parameters are estimated through the numeric maximization of the log-‐likelihood function (Equation 5) derived in Le Gat (2009), using the Nelder-‐Mead nonlinear optimization method. ln ! !, !, ! !, !, !, !, ! =
= −
!
!!!
! !!
!! ln ! +
+ !! ℎ
!! !!
ln ! !! + !
!, !,!!!
!!
! , !! , !!!!
+ !! ln! + !! !! ⊺ ! + ! − 1
!!!
ln !!" +
!!
!!!
⊺
α!!" ! ! !! ! (5) where: m is the number of pipes; ! = [!! … !! ] with !! being the number of recorded failures of pipe ! ; ! = [!! … !! ] with !!" being the age of pipe ! at the ! !! failure; ! = [!! … !! ] with !! being the vector of covariates of pipe ! ; ! = [!! … !! ] with !! being the age of pipe ! at the beginning of observations; ! = [!! … !! ] with !! being the age of pipe ! at the end of observations; ℎ !, !, ! !! , !! , !! = α! !! ! !! ! + ln 1 − exp α! !! ! !! ! − α! !! ! !! ! + exp −α! !! ! !! ! ⊺
⊺
⊺
⊺
Once the LEYP parameters have been estimated, the probability of a given pipe failing ! times during [!, !], knowing it has failed ! times during [!, !], is given by: ! ! ! − ! ! = ! ! ! − ! ! = ! ==
!!!
!!!
! !! + ! + !
⊺
where ! ! = exp α! !! ! ! ! . ! ! −! ! +1
! !! !!
! ! −! !
! ! −! ! +! ! −! ! +1
!
! !! !!!!
(6) The predicted number of failures in a given pipe during [!, !], knowing it has failed ! times during [!, !], is given by Equation (7). E N t − N(s)|N b − N a = j = α!! + j
! ! !! !
! ! !! ! !!
p. 35 / 38
(7) Further reading Le Gat, Y. (2009). Une extension du processus de Yule pour la modélization stochastique des événements récurrents. Application aux défaillances de canalizations d'eau sous pression. Ph.D. thesis, Cemagref Bordeaux, Paristech. Martins, A. (2011). Stochastic models for prediction of pipe failures in water supply systems. MSc thesis, Instituto Superior Técnico, Technical Univ. Lisbon, Portugal Martins, A., Amado, C., Leitão, J.P. (2011). Stochastic models for prediction of pipe failures in water supply systems. (undergoing submission) AWARE-P software documentation: draft version 2012-02-02
p. 36 / 38

software documentation

Transcription

Similar documents

Fernco Flexible Coupling Installation - Fernco

Hymax Couplings 1.5

HEA TER TEMPLA TE PROP END

lfucg follows kytc lead and approves ads and hancor hdpe pipe for

Fernco Flexible Coupling Installation - Fernco

Aviva Stadium

VIDEO SCANNER LASER FOR TV INSPECTION AND

Chart L4: Details (sidewalks, stormwater management)

WaterTurbineAppendix..

The New American Home