2016 water infrastructure conference sun03 - district ... · owwa water distribution committee....
TRANSCRIPT
2016 Water Infrastructure Conference
SUN03 - District Metered Area (DMA)
for Real Loss Management: From Concept to Reality
October 30, 2016 Phoenix, Arizona
Workshop Agenda 8:30 Introduction to Water Loss Management: Auditing, Validation, Water Loss Segregation and
Real Loss Performance Indicators Brian Skeens, CH2M
8:40 Real Loss Management Strategies Overview: With a Focus of Active Leakage Control and
Pressure Management Alain Lalonde
9:05 History of District Metered Areas and Why Use Them
Gary Trachtman, Arcadis 9:30 DMA Design: Using Hydraulic Modeling to Optimize Design: Exercise to Select and Design
DMAs Thomas Walski, Bentley Systems, Inc.
10:15 Break 10:25 DMA Design and Analysis Techniques: Applications to Denver Water
James Uber, Citilogics 10:50 Using DMA Data to Support Water Loss Analytics and Hydraulic Model Calibration
Erick Heath, Innovyze 11:15 Conducting Night Flow Analysis to Calculate Non-Revenue Water
Elio Arniella, Smart Water Analytics 11:40 Case Study Results of Real Loss Reduction using DMAs
Chris Leauber, Water & Wastewater Authority of Wilson County Reid Campbell, Halifax Regional Water Commission
Presenter Information Brian Skeens, CH2M [email protected]
Bio: Mr. Skeens is a Principal Technologist for CH2M. He is currently serving as Deputy Global Service Leader for Conveyance and Storage within CH2M and has over 18 years of experience in water system planning. Brian is an active member of the American Water Works Association (AWWA) Water Loss Committee and contributed to the development of AWWA’s Water Loss Tool and AWWA’s M36 “Water Audits and Loss Control Programs” Manual.
Alain Lalonde [email protected]
Bio: Over his 22-year career in the water industry, Alain Lalonde has help to promote and implement leading edge approaches and methodologies to assess and reduce Non-Revenue Water. A recognized authority in NRW management, Alain has managed some of the largest real loss reduction programs within Canada and has also work internationally in USA, Mexico, Brazil, and Dominican Republic. He is an active member of the AWWA Water Loss Control Committee, IWA Water Loss Specialist Group and OWWA Water Distribution Committee. Alain has presented & published numerous papers on water loss assessment, leakage reduction, pressure management & water metering. He successfully managed the implementation of the first performance based leakage reduction project in Canada valued at $2.2M. Prior to joining Echologics, Mr. Lalonde was a co-founder of one of the leading consulting and contracting companies in NRW management in Canada. He has managed water auditing, leak detection, DMA design & implementation and advanced pressure management programs for several of the largest Cities in Canada including Toronto, Montreal, Calgary, Ottawa and York Region as well as projects in Farmington Hills & Grand Rapids, MI. Alain successfully completed the first water audit and ILI calculation, implemented the first flow modulated pressure management areas, and managed the implementation of over 200 DMAs within Canada. He holds a Bachelor of Applied Science degree in Civil Engineering from the University of Ottawa and is a registered professional engineer.
Gary Trachtman, Arcadis [email protected]
Bio: Principal Environmental Engineer ARCADIS Life Member AWWA, ASCE Member, AWWA Water Loss Control Committee
• Chair, Subcommittee on Water Audit Regulatory Practices • Member, Subcommittee on Strategic Business Planning • Member, Subcommittee on Outreach • Contributor/Editor, M36 Manual on Water Audits and Loss Control Programs (3rd ed., 2009 and
4th ed., 2016)
Secretary, AWWA Customer Metering Practices Committee
• Contributor/Editor, M22 Manual on Sizing Water Service Lines and Meters (3rd ed., 2014)
He has performed water audits for water systems ranging from 5,000 to 400,000 accounts, and has recommended and assisted with implementation of programs for reducing and managing Non-Revenue Water. Participated in WRF Project 2928 – Leakage Management Technologies. Co-author and Presenter on Water Loss Management topics at numerous AWWA technical sessions and Workshops.
Thomas Walski, Bentley Systems, Inc. [email protected]
Bio: Tom Walski is senior product manager for water and wastewater products for Bentley Systems. He has a Ph.D. in environmental and water resources engineering from Vanderbilt University. He has authored several books and several hundred journal papers and conference presentations. He was named one of the 50 icons of the water industry over the past 50 years by Water and Wastes magazine. He has won numerous awards for his work such as the best distribution and plant operation paper in the Journal AWWA on three occasions. He has served as an executive director of the Wyoming Valley Sanitary Authority, engineering manager for Pennsylvania American Water, associate professor of environmental engineering at Wilkes University, an engineer with the Army Corps of Engineers and manager of distribution system operation for the City of Austin, Texas. He co-holds seven patents for hydraulic analysis techniques. He is a registered a professional engineer in two states.
James Uber, Citilogics [email protected]
Bio: Jim Uber is CEO of CitiLogics, and has developed hydraulic and water quality models, and related technologies, for 30 years. Dr. Uber earned a PhD in Environmental Systems Analysis from the University of Illinois in 1988 and was a professor of environmental engineering at the Univeristy of Cincinnati from 1990-2015. He has an extensive background in systems analysis and network modeling, focused on real-time modeling technologies and applications
Erick Heath, Innovyze [email protected]
Bio: Mr. Heath is the Business Director for the Americas and Asia/Pacific for Innovyze. He has over three decades of experience in our industry and is motivated by providing cutting-edge software tools for Utilities and Consultants to implement in our Wet Infrastructure Industry.
Elio Arniella, Smart Water Analytics [email protected]
Bio: Elio has worked in the fields of water and wastewater management for more than 35 years, participating as project director, project manager, or project engineer in more than 200 projects for governmental agencies, international financing institutions, municipalities, and private industry. His experience encompasses all aspects of planning and design for water and wastewater infrastructure projects, ranging from investigations, feasibility studies, design, construction oversight, project financing, information technology, loan procurement, and privatization of infrastructure systems.
He is particularly qualified in regional planning for integral management of water, wastewater and water resources management projects. He has been recognized as an international expert in the area of water efficiency and non-revenue water (NRW) reduction and making water utilities more efficient and sustainable. He has developed innovative methods for monitoring with accuracy and evaluating water loss components in District Metered Areas (DMAs).
Chris Leauber, Water & Wastewater Authority of Wilson County [email protected]
Bio: Chris Leauber is the chair of the AWWA Water Loss Control Committee, Executive Director of the Water & Wastewater Authority of Wilson County, TN and has 35 years of experience in water loss control. He obtained his BS degree from the Pennsylvania State University, managed water loss projects performed on hundreds of utilities prior to joining the Authority and is a certified Grade II Water Distribution Operator. He joined the Authority in 2006 and now manages the overall operations of the Authority and maintains a focus on water loss control.
Reid Campbell, Halifax Regional Water Commission [email protected]
Bio: Reid Campbell is Director of Water Services for Halifax Water. Halifax Water is a water, waste water and storm water utility serving a population of 350,000 in Halifax, Nova Scotia, Canada. Reid joined Halifax Water in 1998 after 10 years of water supply consulting practice.
Reid is responsible for operation of the municipal water supply from source to tap including source water protection, treatment plant operations, transmission and distribution and water quality management. His department is also responsible for development and management of the corporate SCADA system. His responsibilities also include implementation of Halifax Water’s Water Quality Master Plan and Water Loss Control Program.
Reid is a Civil Engineering Graduate of the Technical University of Nova Scotia (now Dalhousie University) and the University of Toronto. He is a member of AWWA and IWA. Reid is a past Vice President of AWWA, a member of the AWWA Water Utility Council.
Workshop Speakers
• Brian Skeens, CH2M
• Alain Lalonde, Echologics
• Gary Trachtman, Arcadis
• Tom Walski, Bentley
• Jim Uber, Citilogics
• Erick Heath, Innovyze
• Elio Arniella, Smart Water Analytics
• Chris Leauber, Water and Wastewater Authority of Wilson County
• Reid Campbell, Halifax Water
Workshop Agenda
• Introduction to Water Loss Management: Auditing, Validation, & Performance Indicators
• Real Loss Management Strategies Overview: with a focus on Active Leakage Control & Pressure Management
• History of District Metered Areas and Why Use Them
• DMA Design: Using Hydraulic Modelling to Optimize Design Exercise to Select and Design DMAs
• Break
• DMA Design and Analysis Techniques: Application to Denver Water
• Using DMA Data to Support Water Loss Analytics and Hydraulic Model Calibration
• Conducting Night Flow Analysis to Calculate Non-Revenue Water
• Case Study Results of Real Loss Reduction using DMAs
District Metered Area (DMA) for Real Loss Management:
From Concept to Reality Prepared by Members of the AWWA EMAC and WLCC
Summary—Why Manage Non-Revenue Water • The US drinking water industry is
facing challenges of resource shortages, aging infrastructure, legal liability, public health and funding needs
• To address these, managing non- revenue water should become a standard business practice
• AWWA is actively promoting the IWA/AWWA Water Audit Method and providing tools for its use
• A number of state/regional agencies are already embracing these methods and applying them
Introduction to Water Loss
Management: Auditing, Validation, & Performance Indicators
Brian M. Skeens, P.E.
CH2M
https://www.youtube.com/watch?v=GMTnUGiSxy0
Introductions
• Name
• Where you work
• Favorite Halloween costume
ustomer Meterin
ystematic Data H
Leakage on
Leakage on Se
Fire Dept Usage
Operat IWA/AWWA Water Balance ional Flushin g
Tools for control include efficient flushing
practices and awareness campaigns
Water
Exported
Non-physical / revenue loss - slowBilledmeters,Revenue
Billed Water Exported
Own billing issues and tAhuethftorized Sources Cost impacts at ‘rCeotnasiul’mrpatitoen.
Authorized Consumption
Water Billed Metered Consumption
ToTooltsal for control include data management, Billed Unmetered Consumption
qSuyastleit
my control policies/practices, & meter
Input testing & repair
Unbilled Authorized
Consumption
Unbilled Metered Consumption
( allow for
Water Supplied
Unbilled Unmetered Consumption
Unauthorized Consumption
Pkhnoywsnical loss – leakage Apparent
Imported Tools for control incWluadteer leakage and
pressure managemLeosnstes
Water C
erorosrts )impacts at ‘production’ rateLosses
Non- Revenue
Water
C
S
g Inaccuracies
andling Errors
Mains
Real rvice Lines
Losses Leakage & Overflows at Storage
IWA/AWWA Standard Water Balance
Water Exported
Billed Water Exported
Own Sources
Authorized Consumption
Billed Authorized
Consumption
Revenue Water Billed Metered Consumption
Total Billed Unmetered Consumption
System Input Unbilled
Authorized Consumption
Unbilled Metered Consumption
( allow for
known errors )
Water Supplied
Unbilled Unmetered Consumption
Apparent Losses
Unauthorized Consumption
Customer Metering Inaccuracies
Water Imported
Water Losses
Non- Revenue
Water
Real Losses
Systematic Data Handling Errors
Leakage on Mains
Leakage on Service Lines
Leakage & Overflows at Storage
M36 4th Edition (2016) Water Audits and Loss Control Programs • Chapter 1 – Introduction: Auditing Water Supply Operations and Controlling Losses
• Chapter 2 – Implementation of Water Loss Control Regulatory Approaches in North America
• Chapter 3 – Conducting the Water Audit
• Chapter 4 – The Occurrence and Impacts of Apparent Losses
• Chapter 5 –Controlling Apparent Losses: Optimized Revenue Captre and Customer Data Integrity
• Chapter 6 – Understanding Real Losses: The Occurrence and Impacts of Leakage
• Chapter 7 – Controlling Real Losses: Leakage and Pressure Management
• Chapter 8 – Planning and Sustaining the Water Loss Control Program
• Chapter 9 – Considerations for Small Systems
• Appendix A – Validating Production Flowmeter Data and the Annual Water Supplied Volume
• Appendix B-E – Blank Forms, Assessingg Water Resources Management, Free Software Tools from AWWA and WRF, Validated Water Audit Data Collection and Analysis
Yes
*** YOUR SCORE IS: 60 out of 100 ***
5 1,000.000 ?
1 100.000
1 100.000
100.000 9 25.000
8 700.000 9 50.000
? 10.313 1.000
10 3.000 3.000
5 7.071
4 5.000
1.00%
0.25% 5.000
? 7 100.0 6 1,000
10
6 60.0
5 $1,000,000
? 7 $3.50 $/1000 gallons (US)
7 $3,000.00
WATER LOSSES: 64.688 MG/Yr
NON-REVENUE WATER
NON-REVENUE WATER: ?
= W ater Losses + Unbilled Metered + Unbilled Unmetered
SYSTEM DATA
Length of mains: +
Number of active AND naci tive service connections: + ?
Service connection density: ?
Are customer meters typically located at the curbstop or property line?
Average length of customer service line: + ?
75.000 MG/Yr
miles
conn./mile main
(length of service line, beyond the property
ft boundary, that is the responsibility of the utility)
Average length of customer service line has been set to zero and a data grading score of 10 has been applied
Average operating pressure: + ? psi
COST DATA
Total annual cost of operating water system: + ?
Customer retail unit cost (applied to Apparent Losses): +
Variable production cost (applied to Real Losses): + ?
$/Year
$/Million gallons Us e Cust om er Ret ai l Uni t Cost to val ue r eal l oss es
WATER AUDIT DATA VALIDITY SCORE:
A weighted scale for the components of consumption and water loss is included in the calculation of the Water Audit Data Validity Score
PRIORITY AREAS FOR ATTENTION:
Based on the information provided, audit accuracy can be improved by addressing the following components:
AWWA Free Water Audit Software:
Reporting Worksheet
Water Audit Report for:
Reporting Year:
W AS v5.0
A mer ican Water Wor ks Associ ati on.
C opyrig ht © 2014, All Rig hts R eser ved.
P lease enter data in the white cells below. W here available, metered values should be used; if metered values are unavailable please estimate a value. Indicate your confidence in the accuracy of the
input data by grading each component (n/a or 1-10) using the drop-down lis t to the left of the input cell. Hover the mouse over the cell to obtain a description of the grades
All volumes to be entered as: MILLION GALLONS (US) PER YEAR
To select the correct data grading for each input, determine the highest grade where the utility meets or exceeds all criteria for that grade and all grades below it. Master Meter Error Adjustments
WATER SUPPLIED <----------- Enter grading in column 'E' and 'J' ----------> Pcnt: Value:
Volume from own sources: + ? MG/Yr + ?
Water imported: + MG/Yr + ?
Water exported: + ? MG/Yr + ?
WATER SUPPLIED: 825.000 MG/Yr
MG/Yr
MG/Yr
MG/Yr
Enter negative % or value for under-registration
Enter positive % or value for over-registration .
AUTHORIZED CONSUMPTION
Billed metered: + ? MG/Yr
Billed unmetered: + ? MG/Yr
Unbilled metered: + ? MG/Yr
Unbilled unmetered: + 9 MG/Yr
Click here: ?
for help using option
buttons below
Pcnt: Value:
1.25% MG/Yr
Default option selected for Unbilled unmetered - a grading of 5 is applied but not displayed
AUTHORIZED CONSUMPTION: ? 760.313 MG/Yr
WATER LOSSES (Water Supplied - Authorized Consumption)
Apparent Losses
Unauthorized consumption: + ?
64.688 MG/Yr
Use buttons to select
percentage of water
supplied
OR
value
MG/Yr
Pcnt:
0.25%
Value:
MG/Yr
Unauthorized consumption volume entered is greater than the recommended default value
Customer metering inaccuracies: + ? MG/Yr
Systematic data handling errors: + ? MG/Yr
MG/Yr
MG/Yr
Apparent Losses: ? 15.071 MG/Yr
Real Losses (Current Annual Real Losses or CARL)
Real Losses = Water Losses - Apparent Losses: ? 49.617 MG/Yr
1: Volu me from own sources
2: Customer metering inaccuracies
3: Total annual cost of operating water system
Data Grading and Validity
• AWWA developed a detailed grading matrix for Water Audit inputs
• Based on the utility’s policies and practices for data collection, data management, data archiving, quality control procedures, and derivation of audit inputs
• Provides a quantitative measure of the reliability
Northern San Leandro Combined Water Sewer Storm Utility District (0007900)
2013 1/2013 - 12/2013
? Cli ck to access defi nit i on
+ Cli ck to add a com ment
Tools for Water Loss Control
• The “M” Series: Manuals of Practice • Guidance Manuals: widely recognized around the world as source
of best practices in water utility operations and management
• AWWA Water Loss Control Committee’s Free Water Audit Software© • Originally released 2006; current Version 5.0 software (2014)
• Water Research Foundation Research Reports • 4372a – Leakage Component Analysis Tool
• Textbooks
• www.awwa.org - type “water loss control” in search box
14
• creates indices for comparison across water systems
15.071
49.617
64.688
Annual cost of Apparent Losses:
9.1%
23.3%
41.29
Real Losses per service connection per day: Operati onal Efficiency:
gallons/connection/day
Real Losses per length of main per day*: gallons/mile/day
N/A
? Infrastructure Leakage Index (ILI) [CARL/UARL]: 3.28
* This performance indicator applies for systems w ith a low service connection density of less than 32 service connections/mile of pipeline
Outputs
• System Attributes
• Performance Indicators • Financial
• Operational Efficiency
AWWA Free Water Audit Software:
System Attributes and Performance Indicators
Water Audit Report for:
Reporting Year:
*** YOUR WATER AUDIT DATA VALIDITY SCORE IS: 60 out of 100 ***
W AS v5.0
American Water Wor ks Association.
Copyright © 2014, All Rights R eserved.
System Attributes:
Apparent Losses:
+ Real Losses:
= Water Losses:
? Unavoidable Annual Real Losses (UARL):
Annual cost of Real Losses:
MG/Yr
MG/ Yr
MG/ Yr
15.13 MG/Yr
$52,747
$148,850 Valued at Variable Production Cost Return to Reporting Worksheet to change this assum piton
Performance Indicators:
Non-revenue water as percent by volume of Water Supplied: Financial:
Non-revenue water as percent by cost of operating sy stem: Real Losses valued at Variable Production Cost
Apparent Losses per service connection per day: gallons/connection/day
N/A
1,359.36
Real Losses per service connection per day per psi pressure: gallons/connection/day/psi
From Above, Real Losses = Current Annual Real Losses (CARL): 49.62 million gallons/year
Northern Sa n Leandro Combin ed Water Sewer Storm Util ity District (0007900)
2013 1/2013 - 12/2013
Key Performance Indicators
• Operational Performance Indicators • defines and quantifies industry standards
• highlights areas of comparison and annual tracking
Apparent Losses per service connection per day:
Real Losses per service connection per day:
Operational Efficiency: Real Losses per length of main per day*:
Real Losses per service connection per day per meter (head) pressure: From Above, Real Losses = Current Annual Real Losses (CARL):
? Infrastructure Leakage Index (ILI) [CARL/UARL]:
Questions?
Pressure
Management
Active
Leakage
Control
Speed and
quality
of repairs
Pipe Materials Management:
selection, installation,
maintenance, renewal,
replacement
Pressure
Management
Current Annual Real Losses
CARL Speed and
Quality of
Repairs
Active
Leakage
Control
Pipeline and
Assets
Management:
Selection,
Installation,
Maintenance,
Renewal,
Replacement
Real Loss Management Strategies
Active Leakage Control & Pressure Management
District Metered Area (DMA) for Real Loss
Management: From Concept to Reality
WIC 2016 Sunday Workshop
Alain Lalonde, P.Eng. – Echologics
October 30, 2016
What is Real Loss Management?
The proactive implementation of techniques,
technologies and approaches aimed at reducing leakage from water distribution
systems to their economic level.
And it goes past simple leak detection….
Types of Water Leaks
Active Leakage Control – “Run-Time”
LEAK DURATION
A L R
TIME Awareness – inception to identification
Location – identification to pin-point
Repair – pin-point to repair
A L R
Examples of leakage “run-time” 75
12 hours reported main break
54,000 gallons
16 Days
5
reported service
connection leak
115,200 gallons
unreported service
on leak
llons
FL
OW
RA
TE
g
pm
g
pm
g
pm
connecti
182.5 Days 525,600 ga 2
A L R
Active Leakage Control Measures • Acoustic / Correlation Surveys
• Noise Logging Surveys
• Night Flow Analysis via District Metered Areas
• Permanent Acoustic Monitoring
• Transmission Main Surveys / Monitoring
• Others: GPR, Tracer Gas, Satellite …
What works best?
Depends on so many factors both technically & cost.
District Metered Areas (DMA)
DMA – Minimum Night Flow 90
Average Zone
Pressure
140
80
120
70
Inflow Rate 100
60
50 80
40 Customer Use 60
30
40
20 Customer Night Use
Minimum 10 Night Leakage on Distribution System
and Customers' Pipework
20
0 0
00 to 01 to 02 to 03 to 04 to 05 to 06 to 07 to 08 to 09 to 10 to 11 to 12 to 13 to14 to 15 to 16 to 17 to 18 to 19 to 20 to 21 to 22 to 23 to 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Time of Day (24 hour clock)
Ave
rag
e Z
on
e P
res
su
re (
me
tre
s)
Infl
ow
Ra
te m
3/h
r
Pressure
Management
Active
Leakage
Control
Speed and quality
of repairs
Pipe Materials Management:
selection, installation,
maintenance, renewal,
replacement
Background Leakage
Before Pressure Management After Pressure Management
Reported Leaks and Bursts Frequency and flow rates of reported leaks reduce
Rate of rise of unreported leakage reduces
Frequency and cost of economic intervention reduces
Background leakage reduces
Unreported
Leakage
Unreported
Leakage
Background
Leakage
Unreported
Leakage
Unreported
Leakage
0,5 1 1,5 2 2,5 3 3,5 4 4,5
TIME (years)
Pressure
Management
Speed and
Quality of
Repairs
Active
Leakage
Control
Current Annual Real Losses
(CARL)
Pipeline and
Assets
Management:
Selection,
Installation,
Maintenance,
Renewal,
Replacement
Pressure Management – Why? Reduction of excess pressures and surges will:
• reduce the flow rates of all existing leaks • best way of reducing ‘background’ (undetectable)
leakage
• reduce the number of new leaks and breaks each year – reducing annual repair costs
• help to defer mains and services replacements
• reduce pressure-dependent consumption • notably external irrigation
As the pipes deteriorate through age (and possibly corrosion), and other
local and seasonal factors, the ‘failure’ pressure gradually reduces until
at some point in time, burst frequency starts to increase significantly
FAILURE RATE COMBINATION OF FACTORS
CAUSES INCREASED
FAILURE RATE
BOOM !!!
Operating range PRESSURE
If the new pipe system experiences surges or variations the
factor of safety is reduced, but the failure rate will remain quite low.
FAILURE RATE NNEEWW PPIIPPEESS,,
SSYYSSTTEEMM WWIITTHH SSUURRGGEESS
Operating range PRESSURE
Consider the situation when new mains and services are laid,
they are designed to withstand existing system pressures
with a large factor of safety, so failure rate is low
FAILURE
RATE NNEEWW PPIIPPEESS,,
GGRRAAVVIITTYY SSYYSSTTEEMM
Operating range PRESSURE
Pressure Management with DMAs • A DMA is a discrete area of a distribution system
ranging in size with one or more metered inputs that is used to calculate and derive the levels of real losses.
• A PMA is a permanently isolated DMA with pressure reducing valves and metering with or without advanced control used to control background and unreported leakage and to assist is reducing break frequencies.
Next, identify if the stabilized pressures at the critical point are
higher than necessary; if so, reduce the excess
to avoid operating system at it’s ‘failure’ pressure.
FAILURE RATE STEP 2: REDUCE
EXCESS PRESSURE
Operating range PRESSURE
The first step in pressure management is to check for the presence
of surges or variations; if they exist, reduce the range and frequency of both
FAILURE
RATE STEP 1: REDUCE SURGES
PRESSURE
Flow Modulated PRV Configuration
Controller
Conventional PRV Configuration
Types of DMA Pressure Management
• Fixed Outlet Control (PRV)
• Time Based Control
• Remote Node Control
• Flow Modulated Control
Thank You!
Alain Lalonde, P.Eng. Director of Business Development (NRW)
Echologics, a Mueller Technologies Company
Real Loss Management - Takeaways
• Key to success – reduce leak “runtime”.
• No “one size fits all” approach.
• Way more tools in the toolbox today!
• DMA & Pressure Management – provides more then just leakage reduction.
Early Implementation – “Step Tests” or Temporary Pitometer Districts
Agenda
• Early Implementation of District Metered Areas
• Features
• Pros and Cons
• Applications
• Other Evolving Approaches
• Why Use DMAs?
History of District Metered Areas and Why Use Them?
SUN03 Workshop
October 30, 2016
Gary B. Trachtman, PE
Early Implementation – “Step Tests” or Temporary Pitometer Districts
Ref. Engr’g & Contracting, March 24, 1915
Early Implementation – “Step Tests” or Temporary Pitometer Districts
Found 6.4
MGD of leakage,
of which
4.5 MGD
was on services
Ref. Engr’g & Contracting, March 24, 1915
Early Implementation – “Step Tests” or Temporary Pitometer Districts
• “…small test pitometer
districts are isolated [from 11 pm to 5 am] and the
controlling valve of each
section of main is
operated, the drop in rate of flow at the pitometer
being noted.”
• “…the aquaphone
inspection is now made at a time when the quantity
of water flowing is known,
and not several days after
the night test, as formerly,
and the curb stops are operated at night, when
the flow is steady, making
for greater accuracy in the • “…[customer] meters are examined at night…” • “pitometers are placed on smallest mains available [to
minimize need for opening a fire hydrant]…”
results and causing less inconvenience to
householders.”
Features of DMAs
• Mandated in UK and Wales since mid-1980s, used extensively globally, increasingly used in North America
• Features • Isolated area with roughly 1k to 3k service connections
• Preferably 1 entry point with inflow small enough to quantify individual leakage events
• Preferably simultaneous measurement of customer demand
DMA Design
Options
United Kingdom Water Industry Research (UKWIR). 1999. A Manual of DMA Practice. Report Ref. No. 99/WM/08/23 United Kingdom
Pros and Cons of DMAs
Pros
Deviations from normal flows and pressures are quickly evident
Can reduce response time for leak repair
Helps in prioritizing leak pinpointing efforts when leakage volume exceeds economic level
Enables advanced analysis of customer consumption and leakage patterns and their quantities
Can control background leakage when used in conjunction with proactive pressure management
Can extend life of mains by managing pressure
Can support water conservation efforts by reducing pressure- dependent demands
Pros and Cons of DMAs
Cons
Requires varying degree of capital investment, depending on configuration and operational status of valves, appurtenances, and data communication capability
Requires careful planning, design and incremental start-up to maintain adequate hydraulic (domestic and fire flows) and water quality performance
Adds to facility maintenance requirements
Need to re-program autoflushers to avoid minimum demand periods
DMA Application in Philadelphia
DMA 5 Pilot Program (2005-2010)
• 1,266 connections evaluated, 12.6 mi metallic pipe (avg age 52 yrs), 117 fire hydrants, 382 valves
• Leakage reduced by 1.19 MGD
• Potential reduced frequency of leak survey
• $380,000 capital cost, $44,000 operating expense (annual revenue loss, leak survey, O&M)
• Net annual savings approx. $55,000; payback 6.4 yrs
• Adding fixed-network AMR capability to obtain more frequent customer meter readings
Ref. JAWWA, Kunkel and Sturm, Vol. 103 No.2 (2011)
System-wide Application of DMAs – Halifax Water
• 65 contiguous DMAs with 110 pressure-control and meter chambers
• Advanced Pressure Management in some DMAs
• Reduced average annual breaks from 25 to 12
• Background leakage reduced without significant reduction in consumption for residential and commercial customers
• Resolved challenges involving control complexities
Ref. AWWA Opflow, Dec. 2011; WRF 2928 Leakage Management Technologies (2005)
Applications -Temporary DMAs
• Leak Survey Frequency Analysis
• Allocate resources
effectively
• Consider all relevant valuations and costs when determining: • locations and
frequency of leak survey
• benefit:cost of alternative interventions such as meter change- out, leak repair, or main replacement
BWWB - Some DMA Considerations
• Existing PRV Chamber with Pressure Monitoring, 26 miles from COR
• Acquired system with small diameter PVC, numerous leaks
• Modified chamber to accommodate flow-modulated pressure control valve
Secondary Point Elev 526+/- Entry Point
Elev 520+/-
Critical Point
Elev 580+/-
BWWB - Trafford DMA
Range of Elevation Served = 150 feet +/-
BWWB Trafford DMA – General Location
• 435 mostly residential connections
• 10.4 mi mostly 2”-6” PVC, 8” DI primary feed
• Monthly leak surveys (Pre-DMA)
• 28 Permalogs, Annual leak surveys (Post-DMA)
Ref. King, et al, AWWA DSS 2011
( m g d )
flow
(gp
m)
inle
t pre
ssure
(psi) o
utle
t pre
ssure
(psi)
30
0
Lea
k Even
t
25
0
20
0
150
10
0
50
0
date &
time
BWWB - Results – Trafford DMA
• Found 50 gpm from each of 3 leaks immediately
after implementation
• Reduced reported leakage by reducing running time by at least 30 days (2.2 MG per 50-gpm leak, worth $812 per leak or a total of $7,310 over 3 years)
• Reduced frequency of leak survey from monthly to annually, saving $10,560 in labor cost over 3 years
• Payback period on initial $75k investment is $75,000 ÷ ($10,560 + $7,310) or roughly 12 years
Other Evolving Approaches – DMA or non-DMA
Cloud-based analysis (SaaS, NaaS) of acoustic sensor data:
Utilization of AMI data transmission capability improves monitoring of customer meter performance
Smartphone-based display of pressure comparison, correlation results and anomaly alert messaging
Increased and more timely exposure of hidden leaks, fewer false positives, reduced repair costs
Integration of leak events with Work Order Management Systems
Improved customer service
Other Evolving Approaches – To DMA or Not to DMA? That is The Question
Dynamic Sectorization (non-DMA)
One flowmeter plus hydraulically actuated isolation valves (lower cost to implement)
Sector temporarily isolated (typically at night)
Automated calculation of water balance and leakage
Can be applied in portion of a system that employs permanent DMAs elsewhere
e.g., consider applying to existing temporary Pitometer Districts
THANK YOU!!
Together we can do a world of good.
Gary B. Trachtman, PE
Principal Environmental Engineer
M 205 910 8083
So… Why Consider DMAs?
• Minimize background and unreported leakage, and
inefficient use of water resource
• Reduce operating costs associated with excess leak survey
frequency in remote areas of the distribution system
• Minimize initial expenditure by utilizing existing
infrastructure, where possible
• Minimize collateral expenses from pipe failure
• Maintain level of service (quantity/quality), incl. fire
protection, with more timely response to leakage events
bility
Objectives
• Measure flow for small area
• Pressure management
• Minimal adverse impact on pressure, fire flow, relia
• Minimize costs or metering, valves, vaults
• Similar DMA size
Overview
• Design Considerations
• Hydraulic Design – Modeling
• Other Impacts
• Case Study
• Metering Technology
• Workshop
DMA Design: Hydraulic Modeling and other Considerations
Tom Walski, Ph.D., P.E.
Bentley Systems
Hydraulics
• Identify normally closed isolation valves
• Identify meter location(s)
• Pressure control ?
• Pressure impacts
• Fire flows
• “Sleeper” feeds
DMA Design Considerations
• Isolation
• Flow metering
• Pressure control
• Fire flow impacts
• Water quality
• Flushing
• Reliability
Pressure Control
No DMA DMA
Valve Location
X
Transmission
Distribution
X X X
X
Closed valve
Meter
Sleeper
Don’t place closed valves on Transmission mains
Pressure Management
• Not required
• Can reduce leakage
• Can reduce pipe breaks
• Can impair service
Fire Flow Comparison
No DMA DMA
Fire Flow
No DMA
Fire Flow
With DMA’s
ent
Temporary DMA’s
• Only close boundary valves at night
• Leave system open during high demand times
• Large cost for motorized valves and controls
Water Quality Impacts
• Isolation of contamination – passive containm
• More controlled flushing
• Stale water at DMA boundaries
Flushing Impacts
Types of Sub-Systems to be Isolated
• Pressure zone – elevation based
• DMA – metering, non-revenue
• Distribution block – water quality
• Segment – shut down isolation
Modularity Index - Maximize
Cuts
Module Similarity
Research on Automated DMA Design
• Usually based on graph theory
• Multi objective • Cost
• Minimal impact
• Reduce background leakage
• Diao – edge connectedness
• Giustolisi et al. – modularity index, conceptual cuts
• Grayman et al. – distribution blocks
Wilkes-Barre/Scranton System
• 135,000 customers
• Luzerne & Lackawanna Counties, Pa.
• 70 miles x 10 miles
• 50 MGD average day use
• 10 WTPs
Wilkes-Barre/
Scranton
ESTABLISHING A SYSTEM SUBMETERING PROJECT
Tom Walski
Dave Kaufman, PAWC
Tony Gangemi, PAWC
Bill Malos, PAWC
Where to Locate Meters?
• Boundaries of Sub-metered Zones
• Pump Stations
• PRVs
• Key flow splits
Additional Metering Needed
• Guide loss reduction program
• Notice trends in loss
• Pinpoint areas with high water loss
• Supplement SCADA system
Wilkes-Barre/Scranton System
Fallbrook Forest City
Chinchilla
Brownell
Huntsville Lake Scranton
Ceasetown Nesbitt
Watres
Crystal Lake
Mag Meter
Turbine Meter
Design Issues
• Type of Meter
• Power Supply
• Totalizing or Rate
• Link to SCADA
Differential Pressure PRV
Venturi Meter
Mag
Turbine
Venturi
Meter Comparison
Flow
Cost
Turbine / PRV
PRV/Turbine
Position Sensor
Pressure Sensor
Ancillary Issues
• Installation of new vaults
• Delivery of power to site
• Installation of sump pump
• Installation of wiring
• Installation of SCADA RTU
• Installation of pressure transducers
• Cellular/radio/leased/satellite
• Programming of SCADA computers
Valve: FLAT ROAD #2
Discharge Varying Time
6000.0
5500.0
Low
High
Typical
5000.0
4500.0
4000.0
3500.0
3000.0
2500.0
2000.0
1500.0
1000.0
0.0 5.0 10.0 15.0 20.0 25.0 30.0 35.0 40.0 45.0 50.0
Time
(hr)
Typical PRV
Dis
ch
arg
e
(g p
m)
Typical RTU
Power
Readouts
Multiple DMA’s
4 – vaults, PRVs, RTU’s, Radios 1 building, 1 PRV, 1 RTU, 1 Radio
Un-burying PRV’s and Meters
Typical Data
Church St.
900 90
800
80
700
600 70
500
60 Flow
Pressure 400
300 50
200
40
100
0 30
0 1 2 3 4 5 6 7
Time, days
Flo
w,
gp
m
Pre
ssu
re,
ps
i
Pros and Cons
Flow accountability
Leak reduction
Water quality
Flushing
Fire flow
Pressure
Reliability Cost
• Results
n-revenue water calculation Better no
• Guide leak detection efforts
• Improved SCADA
• Better troubleshooting
• New PRVs
90
85
80
75
70
65
-100 0 100 200 300 400 500 600
Flow, gpm
Pressure/Flow Relationship
Pre
ssu
re,
ps
i
Thank you – Merci – Gracie – Gracias –
Danke - Dziękuję - Dziękujemy - 谢谢
Laying out District Metered Areas
The attached drawings are for section of land to be developed in one US land survey section (one mile square). It is bounded by larger transmission mains in the arterial roads with no customer taps, and is divided into land development projects designated by the letters. The number next to the letters indicate the number of equivalent dwelling units (EDU) in each development. There are only 6 and 8 in. pipes within the developments and there are no elevated tanks or pumps.
Most are average residential areas but D is an exclusive neighborhood with large lots, E is a golf course and F contains apartment buildings and condos. The blue symbols are isolation valves which are normally open. Entrance roads to the developments are shown in green.
1. In the first figure, the topography is relatively flat so all of this section can be considered in the same pressure zone. You job is to determine where to put flow meters and closed valves to define the DMA’s.
2. In the second figure, there are two pressure zones and the master plan says that elevations above 550 are to be served by the high (750 ft HGL) pressure zone which could have heads down to 730 ft. Elevations below 550 ft are to be served by the low (650 ft HGL) pressure zone which is the setting on the PRV. The contours are shown on the drawing in orange. There is one PRV on the transmission mains connecting the two zones and one closed valve in red.
3. In the third figure, you are now working with an existing water system that is fully built out. The pressure zones have long since been established. There is an existing street that runs north to south through the section and has a 12 in. pipe in it with a PRV. Individual customers along the 12 in. are served out of that pipe, such that for example, customers between A and B and G and F are served out of the same pipe.
Work on one figure at a time and we will discuss each before moving on.
24 in. 570
High Zone 550
PRV A - 60 B - 110 Low Zone
12 in.
530
D - 20
C - 80 1 mile
16 in. E - 10 16 in.
510
F - 200 G - 80
12 in.
24 in. 570
High Zone 550
PRV A - 60 B - 110 Low Zone
530
D - 20
C - 80 1 mile
16 in. E - 10 16 in.
510
F - 200 G - 80
12 in.
24 in.
A - 60 B - 110
D - 20
C - 80 1 mile
16 in. E - 10 16 in.
F - 200 G - 80
12 in.
Motivation
▪ The DMA is a general topological network concept — a
minimal region of the system within which a
mathematical flow balance can be constructed from
available data streams
▪ By separating the network into subregions within which
the demand variation is known, the DMA relates
system hydraulic supply and demand, in the same way
that the watershed relates rainfall and tributary flows
▪ By this fundamental definition, there is no size
restriction or expectation for a DMA, except what is
necessary for a particular use
▪ Network modeling provides a natural motivation for the
DMA concept — the smaller the DMA regions, the
more precision in the assignment of demand variability
Outline
▪ Motivation
▪ DMA discovery - what DMAs do you already have?
▪ DMA design - new meters and/or valve closures
▪ Denver Water application
DMA Design and Analysis Techniques: Application to
Denver Water
Jim Uber, Sam Hatchett — CitiLogics
Myron Nealey, Cindy Marshall, Jarrod Loran — Denver Water
Jerry Edwards, Nathan Roberts, BHI
Bryon Wood, Robin Hegedus — HDR
A network might already be divided
into one or several DMAs, waiting to
be “discovered” using automated
techniques similar to those
employed for pressure zone
discovery
DMAs are collections of nodes
delimited by closed pipes or
pipes with measured flows Q
L V Q
DMA
Q
Automated DMA discovery
Automated DMA design
Given the flow sensor and
closed valve locations,
software can determine what
we need to know about
existing DMAs:
geometry
boundary flows and
orientation
tank storage
Unlike mapping of pressure zones,
you will have to map the locations of
existing flow sensors
Denver Water Application — Existing DMAs
Graph Partitioning for DMA Design
▪ Useful to have balanced DMA sizes for diagnostic
purposes - size measured by pipeline length or demand
▪ Implementing DMAs comes with a cost
▪ New flow instrumentation in new vaults ($$$)
▪ New flow instrumentation in existing vaults ($$)
▪ Closing valves ($)
▪ Sounds like a partitioning of the network graph: Node
weights are either demand or attributed pipeline length,
and should be balanced across DMAs; Edge costs are the
cost to isolate or measure, and should be minimized Public
domain graph partitioning algorithm by Karypis (METIS) is
efficient and effective
Efficient Graph Partitioning Algorithms Can be
Applied for DMA Design
▪ Developed for executing tasks in parallel on multiple processors in order to
balance the computational load and minimize communications between processors
▪ The computational task is described as a weighted graph G = (N, E, WN, WE)
▪ Nodes (N) represent independent tasks; Node weights (WN) are computational costs of each task
▪ Edges (E) represent communication required between the independent
tasks; Edge weights (WE) are the amount of data that needs to be
transferred between tasks in order to complete the overall job
▪ Partitioning G means dividing the nodes N into the union of P disjoint
collections, where one processor will handle all the jobs in a collection.
▪ Partitioning is to respect the goals: 1) the total load is balanced among the
processors, or the sum of the node weights in each partition is
approximately equal; 2) the total communications volume between different processors is minimized, or the sum of the weights on edges
connecting nodes in different partitions is a minimum
Denver Water Application —
DMA Partitioning Analysis Using Additional Flow Measures
Existing flow
measures, known
closed pipes and
valves — Two
unbalanced DMAs
77 new measures
25 new measures
14 new measures
Summary
You may have DMAs already that only need to be
discovered and leveraged
Graph partition algorithms are computationally efficient (Less
than one minute per partition; 300,000+ nodes/pipes) and
can be leveraged for designing efficient monitoring schemes
to further subdivide the network into balanced DMAs
A modern data integration environment DMAs to be
efficiently identified and used for demand analysis and data-
fused (“real-time”) network simulation
Software Demo
(given time)
DMA Data for Water Loss Analytics
A District Metered Area is a mini-pressure zone
• Short-term (or permanent) flow meters installed at key locations
• Measure: • All flows into each DMA
• All flows out of each DMA
• All changes in storage tanks levels (converted to flow rates)
• Customer Demand Data • AMR/AMI data is best (15 min or 1 hour data preferred)
• If no AMI, monthly billing data is still useful
• Diurnal Usage Pattern = Σ Inflows – Σ Outflows ± ∆ Storage
• NRW = Diurnal Usage – Water Consumption
DMA NRW • Which pressure zones (or DMAs) have the highest losses?
calcs • How can I get real-time Non-Revenue Water (NRW) numbers?
model
• Have pump curves been validated?
hydraulic modeling challenges
updating • Are diurnal curves accurately defined?
• How calibrated is the model? calibration/ • What’s the confidence level validation on model results?
SCADA bottleneck
operational • Does the model reflect day-to- modeling day operations?
• Do operators trust model results?
• How many steps to retrieve data from SCADA?
• Does the model always match SCADA?
Using DMA Data to Support Water Loss Analytics and Hydraulic Model Calibration
SUN03 Workshop
October 30, 2016
Erick Heath, PE
Diurnal Curve & NRW Calculations • Match model IDs to field • Zone data calculated at any user
measurements/SCADA tags defined intervals (daily, monthly, etc.)
• Hydraulic model relationships • Diurnal curves and NRW calculated & drive zone configurations imported to hydraulic model
DMAs vs. Pressure Zones (small example)
• Model below has: • Four pressure zones (DMAs)
• One water source (WTP)
• Five pumps
• Four control valves
• Three tanks
Hydraulic Model Calibration
The most technically challenging aspect of the hydraulic model is:
Reporting &
results
presentation Other Water quali2t%y
analysis
9%
5% Building/updating
the model
18% The Frequency of Hydraulic Model Calibration
35
30 Developing &
analyzing
planning &
operational
scenarios
15%
Integration with
GIS
9% 25
20
Calibrating the
model 15 42%
10
5
0
< 1 yr 1-2 yrs 2-4 yrs > 5 yrs
*Oct 2014 AWWA Journal - Committee Report: Trends in water distribution system modeling. This report discusses the results of the EMAC 2013 survey, the past and current modeling issues challenging utilities, and trends that will shape distribution network modeling. Data above is based on 209 survey responses.
Pe
rce
nta
ge
of
Re
spo
nse
s
Typical Modeling Approach
A disproportionate amount of resources are typically applied to the building, development, and calibration of models compared to the analysis of those same models.
Model Build/Update Model Development Validation/Calibration
Boundary Conditions Network of pipes and nodes
Controls
Elevations Operational Patterns
Tanks, Pumps, and Valves
Demand Scaling, NRW
Field Conditions
Pump Validation
Facility Attributes Results Sharing
Pump Curves
Assign Demands
C-Factors
Controls
Operational Patterns
Demand Scaling, NRW
System-wide ADD, MDD, PHD
Diurnals
SCADA Integration
SCADA
real-time data server
Historian
secure read only access to SCADA
Server
Auto reads data from SCADA
and connects directly to
hydraulic model
Model
auto updated from GIS and SCADA (via
SCADAWatch)
Access to SCADA Data AND latest
hydraulic model results on any
device at any time!
GIS
continually updated Demands
updated from AMR/AMI (via SCADAWatch)
SCADA Integrated Modeling Approach
A direct connection between SCADA data and the model streamlines model development and calibration aspects of modeling and makes calibration a continual process.
Model Build/Update Model Development Validation/Calibration
Network of pipes and nodes Pump Curves
Assign Demands
Elevations C-Factors
Tanks, Pumps, and Valves
Facility Attributes
Controls
Operational Patterns
Demand Scaling, NRW
System-wide ADD, MDD, PHD
Diurnals
Boundary Conditions
Controls
Operational Patterns
Demand Scaling, NRW
Field Conditions
Pump Curve Validation
Diurnals
Results Sharing
And much more…
Diurnal Curve Creation
Old School
=PI()*((K$35/2)^2)*(J44-J43)*7.48
Benefits of SCADA & Model Connection
A permanent connection between field sensors (or SCADA data) and a hydraulic model for calibration
• Generation of diurnal usage (demand) curves per pressure zone
• Determination of operational pump curves
• Validation of pump and valve controls
• Utilization of continuously calibrated operational models • Hydraulics
• Water Quality
• Energy
• Carbon Footprint
• Other…
• Automatically rerun a model based on current SCADA information every XX minutes
Diurnal Curve Creation
Validating Pump Curves
Validating Pump and
Valve Controls
Calibration
SCADA: Friend or Foe
Validating Pump Curves
Model Integrated with SCADA
• Auto-generate real-time and historical pump curves
• Determine individual pump curves from multi-pump pump stations
Validating Pump Curves Old School
Diurnal Curve Creation
Model Integrated with SCADA
Validating Pump and Valve Controls
Model integrated with SCADA
Validating Pump and Valve Controls
Model integrated with SCADA
Validating Pump and Valve Controls
Old School
Summer vs. Winter controls?
SCADA Integration Sounds Difficult
Just a Few Steps Involved
• Match SCADA Tags to Model IDs
• Define Inflows, Outflows, and Tank Levels for each Pressure Zone (via drag/drop
interface)
• Connect to Billing Database (AMR/AMI/Monthly Data Reads, etc.)
• Configure Pump Station sensors (flow vs. head [convert from pressure])
• Select date/time of interest (summer, winter, historic pipe break, etc.)
• Run model!
Let’s Look at a Real User Case Study…
Calibration Model integrated with SCADA
Calibration Old School
=IFERROR(INDEX('Field Data'!$A$1:$EZ$10000,MATCH($B53,'Field
Data'!$A:$A,0),MATCH(G$1,'Field Data'!$A$1:$EZ$1,0)),"")
Yorba Linda Water District, CA
YLWD Diurnal Curves Jan & July Ave ~ 0.4 MG
YLWD System Mass Balance
GW Well Inflows
Tank Storage Volume Changes
Orange County
MWD Inflows
Single Click Diurnal Curves
Ave ~ 0.9 MG
Yorba Linda Water District, CA
Non-Revenue Water Hidden Hills PZ
Hidden Hills Pressure Zone Mass Balance
• NRW – yes
• Energy Numbers
• Amperage
• KWHrs • Other
• Carbon Footprint
Hidden Hills PZ Diurnal Curves: Jan & July
Ave ~ 5,500 gph
• Jan = 31 days • Diurnal for each of 24
hours across all days in Jan (and later in
July – also 31 days)
Ave ~ 15,000 gph
Thanks for attending!
Erick Heath, P.E. Vice President 626.568.6855 [email protected]
Utilities w/SCADA Integrated Models
• Boulder, CO
• Yorba Linda W.D., CA
• Rancho California W.D., CA
• Berkeley County W&S, SC
• Castaic Lake W.A., CA
• Golden State Water Co., CA
• Regional Water Authority, CT
• … and Others…
Current Pilots
EBMUD, CA
SFPUC, CA
Cal Water, CA
Benefits of SCADA Integrated Modeling
Streamlined modeling approach
• Enhances model update process
• Makes calibration and validation magnitudes easier
• Model is always calibrated
• Provides platform for real-time modeling
Advanced modeling applications are now a reality
• Normal and emergency response using *current* model results
• Groundwater well management
• Energy use minimization
• Non-Revenue/Water Loss analysis
• Real-time forecasting of model results
Increased ROI on model, SCADA, and GIS investments
Questions & Answers
Objectives Present a quick method to analyze Non-Revenue Water
(NRW) in DMAs
Show how continuous monitoring of night flows into water supply zones or DMAs is an important operational tool for identifying water loss within a network
Discuss important considerations for monitoring night flows
Show how to enhance the validity of MNF analysis by monitoring consumption
Conducting Night Flow Analysis to
Calculate Non-Revenue Water
WIC16 Workshop SUN03
District Metered Area (DMA) for Real Loss Management:
From Concept to Reality October 30, 2016
Elio F. Arniella, P.E.
Background Information
Non-Revenue Water (NRW) is the difference
between the amount of water flowing into the system minus the total amount of water billed to the customers
The total water losses in a Service Area or DMA
are defined by the Minimum Night Flow (MNF), Customer Night Consumption, and Flow supplied to a DMA
Minimum
Night Flow
Minimum Night
Consumption
Water
Losses/Leakage
Night Flow Analysis
Background Information, Cont.
AMI/AMR metering provide new opportunities for determining accurately NRW in a DMA using MNF analysis
Management Information Systems need to be updated to handle AMR/AMI
Customer meters need to be sensitive to typical night flow volumes
Methodology
Monitoring water supplied to a DMA with close or open boundaries
Monitor input supply
Monitor customer consumption
Monitor night consumption
Compare the NRW obtained by water balance with the NRW volume obtained by the MNF methodology
Develop a night consumption pattern of the DMA
Estimate NRW with just the DMA supply data
Works best in areas with mostly residential customers and small commercial accounts without nighttime activities
DMA Case Studies
DMA: 2,900 Customers
AMI Customer metering
59 miles of water mains
Mostly residential use
Water purchased in bulk
Two entry points
Important considerations
Careful consideration must be given to the
following issues:
Sizing the meter appropriately for accuracy at low flows
Locating the meter(s) so that all inflows to a
zone are captured, including storage tanks
(filling and emptying dynamics)
Establish night consumption patterns by monitoring customer meters.
Important considerations cont. Night consumption monitoring should include meters
that can read, at least, in 1 gallon increment and with capabilities to store data in 15 minute to 1 hour time step
Storage tanks inside the DMA affect the supply patterns. Storage tank inflow, outflow and level must be monitored with high level of precision in order to obtain accurate supply patterns.
A calibrated hydraulic models can help to establish the consumption patterns
Monitor of possible reverse flows, specially in DMAs with storage tanks. Subtract the reverse flows from the supply to the DMA for the water balance
AMI Customer Consumption: 1 Hr. Data
Insertion Probes Water Supply : 15 min. Data
Combined Flow for Two Insertion Probes Water Supply : 15 min. Data
Sup
ply
, G
PM
Night Consumption Analysis for
Several US Utilities Average
Source Author MNC, GPH
AWWA Research Foundation AWWA 1.05
Georgia Utility 1 SWA 1.95
Georgia Utility 2 SWA 1.42
Georgia Utility 3 SWA 1.06
Georgia Utility 4 SWA 2.06
Ohio Utility 1 SWA 1.41
Night Consumption Analysis
(10 days and >200 customers)
DMA Customer Consumption
Manual 15 min. Data for 120 customers
71.3 GPM = 1.41 GPH
Questions?
Thank you !
Elio F. Arniella, P.E.
NRW Minimum
Night Flow
Minimum Night
Consumption
Conclusions
Measuring consumption in a DMA provides a quick method to analyze Non-Revenue Water (NRW) in DMAs
It is important to have customer meters that read in 1 gal. increments
Typical night consumption in predominantly residential areas ranges from 1 to 2 gallons per hour
Night Consumption Analysis
Average for 31 days for 40 customers
www.hrwc.ca Slide 3
Dartmouth Central
WATER LOSS CONTROL AT HALIFAX WATER
• Adopted IWA/AWWA Methodology in 2000.
• Reduced ILI from 9.0 to 2.4
• System Inputs reduced from 168 to 130 MLD.
• 45 pressure zones, 75 DMA’s, 1 PMA
Pressure Management – Utility Case Study
HALIFAX WATER’S EXPERIENCE WITH
PRESSURE MANGAGEMENT
Reid Campbell
Halifax Water October 30, 2016
DARTMOUTH CENTRAL PMA BEFORE AND AFTER
Avg Annual 2002/03 – 2004/05
Main Leaks = 23
Pub. Service Leaks = 4
Priv. Service Leaks = 5
2005/2006
Main Leaks = 12
Pub. Service Leaks = 2
Priv. Service Leaks = 3
Since 2005/2006: 16 /year
Pressure Management
UARL
Potentially
Recoverable Annual
Volume of Real Losses
PRESSURE MANAGEMENT AT HALIFAX WATER
• Based on several years of flow modulated pressure control:
• Breaks are measurably reduced. • Dependent on system characteristics.
• At the time thought that it had important but limited application across the utility.
• Dartmouth Central PMA/DMA has operated since 2005-06. • Changed to solenoid control in 2013-14.
• Based on recent data analysis opportunity for pressure management is much greater.
DARTMOUTH CENTRAL PMA
• Flow modulated pressure control
• Controller reduces system pressure as demands decrease
Pressure Flow
CASE STUDY: COLLINS PARK SYSTEM
• Collins Park is a small system owned by Halifax Water: • Commissioned 1988. • Average system pressure: 70 psi. • 75 residential customers. • Ductile iron distribution system (2 km’s). • Direct pressure provided by pumps at the WTP. • No specific break/leak history.
• In response to new Provincial Water Treatment standards, a new treatment plant was constructed. • Commissioned in June 2010.
COLLINS PARK SYSTEM
• 5 Breaks in two months after commissioning: • Pressure increased by 15 psi
• Pressure surges detected by logging
• Quality of Repairs.
PROACTIVE PRESSURE REDUCTION
Leaks and Breaks 2005-06 to 2014-15
• Pressure reduced slightly in some zones. • 6 zones with
reduced pressure.
• 2-5 psi reduction.
Pressure spikes in the
distribution system
captured on SCADA as
reservoir altitude valve
closes too quickly
Main failure
correlates to
pressure spikes
NEW STRATEGIES FAILURE ANALYSIS
WHERE ARE WE GOING NEXT ? BREAKS Vs. TIME OF DAY
NEW FINDING
• In 2012 we did a statistical analysis of all watermain
breaks.
• Key findings: • 60% of all breaks happen between midnight and 6
am. • Similar occurrence for all break types. • Break clusters.
August 20, 2015
• On August 20, 2015, we had 6 breaks in a single zone over a 6 hour period.
During low night
flows, system
pressure creeps up
60 kpa as a result of
a malfunctioning
PRV.
NEW STRATEGIES CREEPING PRV
NEW STRATEGIES
• Night time pressure creep in PRV fed zones. • PRV’s less precise in low end of flow range.
• Due to water loss control: • PRV’s night time operation in lower end of design range. • Less leakage to provide base line flow through prv. • Less leakage to dampen HGL
• Need for improved maintenance.
• Analysis of individual breaks will identify high priority prv’s for maintenance or re-configuration.
UTE
CK BL
V
D
V E R
LA N E
Questions or Comments?
Location of Breaks
Bedford South Booster
Peakview PRV
SM
ITH
S R
D
Water Audit Performance Indicators used for Real Loss Reduction
District Metered Areas (DMAs)
• Presently 336 miles of distribution main
• 15 system input meters
• 5 ground storage tanks
• Established DMAs based on input meters, storage tanks & pump stations • 21 total DMAs
• Convert tank level changes to flow rates • Minimum night flow (MNF) = 1:30 – 3:30 AM
• Legitimate nighttime consumption (LNC) • 1.5 gals/conn/hr
DMAs – Utility Case Study
Chris Leauber
Water & Wastewater Authority of Wilson County, TN
Minimum Night Flow (MNF) Measurements & Calculations
• UARL 60.78 MG/Yr
(0.36 gpm/mi)
• TF DMA = 56 miles
main x 0.36 gpm/mi =
20 gpm
• LNC = 33 gpm
• Total = 53 gpm
• If MNF > 53 gpm intervene
• TF DMA MNF=99 gpm
• TF DMA LNC=33 gpm
• TF DMA Real Losses
(Leakage) = 99 gpm –
33 gpm = 66 gpm
• TF DMA miles main=56
• 66 gpm/56 miles = 1.2
gpm/mi >0.36 gpm/mi
Systemwide DMAs Trousdale Ferry (TF) DMA
Target Setting
• Technical expertise in-house
• Cost of water supply very high • Presently $2.59/1000 gals
• Set goal to achieve & maintain ILI = 1.0 • Should NOT set ILI goal without an economic analysis
• ILI = 1.0 is UARL = 60.78 MG/Yr • 321 miles main in 2009 = 519 gals/day/mi = 0.36 gpm/mi
Step Testing
• TF DMA SW MNF = 38 gpm
• TF DMA SW LNC = 8 gpm
• TF DMA SW Real Losses (Leakage) = 38 gpm – 8 gpm = 30 gpm
• TF DMA SW miles main = 14
• 30 gpm/14 miles = 2.1 gpm/mi > 0.36 gpm/mi
• Note: Water is passing through valve, actual leak rate is greater
Step Testing via Valve Isolations into Smaller Areas &
Calculate Leakage per Mile of Main
Minimum Night Flow (MNF) Measurements & Calculations
Transducers Mounted on Tank Main Line Outside Mounted Flow Display
Portable Mount Flow Meter
Prior & After Repairs
78 gpm Leak = 41 MG/Yr @ $2.59/1000 gal = $106,000/Yr
Leak Sounding Leak Noise only Audible by Ground Microphone at Leak Location
Additional Step Testing within TF DMA SW
• Leakage isolated 3:30 AM to 4,900’ area
• Step tested 56 miles in < 3 hours
• No water surfacing
• 6” PVC main located under soil conditions, 20’ off road , 50 psi
• No low pressures complaints
• Not detectable by direct contact sounding on system valves, hydrants & services
Chris Leauber
Executive Director
Water & Wastewater Authority of Wilson County
P.O. Box 545
680 Maddox Simpson Parkway
Lebanon, TN 37088
E-mail: [email protected]
Tel: 615-449-2951
Maintaining ILI @ Technical Minimum. 2012 drop due to Average System Pressure Estimate Increased for 60 psi to 80 psi, Compiler Software
Water & Wastewater Authority Wilson Co Program Results
• 2014 - Maintaining Technical Minimum, ILI = 0.88
• Real Losses = $186,000/yr.
• TN median ILI 2.17 (Data from Sturm et al. 2015. ©Water Research Foundation)
• If WWAWC 2.17 = $457,000/yr. Real Losses
• $457,000 - $186,000 = $271,000/yr. difference
• $210,000 capital cost investment to maintain ILI, 10 yr. deprecation = $21,000/yr.
• $271,000 - $21,000 = $250,000/yr. net benefit
• $250,000/6,000 customers = $42/cust/yr. net value