03b. ERDC-CERL TR-20-1.pdf

ER

D

C/

CE

RL

T

R-

-1

Guidelines for Minimizing Potable Water Loss Due to Water Distribution System Flushing

Co ns tr uc tio n

En gi ne er in g

R es ea rc h La bo ra to ry

Elisabeth M. Jenicek, Victoria E. Heath, Noah W. Garfinkle, and Andy Hur

January 2020

Approved for public release; distribution unlimited.

The U.S. Army Engineer Research and Development Center (ERDC) solves the nation’s toughest engineering and environmental challenges. ERDC develops innovative solutions in civil and military engineering, geospatial sciences, water resources, and environmental sciences for the Army, the Department of Defense, civilian agencies, and our nation’s public good. Find out more at www.erdc.usace.Army.mil.

To search for other technical reports published by ERDC, visit the ERDC online library at http://acwc.sdp.sirsi.net/client/default.

http://www.erdc.usace.army.mil/ http://acwc.sdp.sirsi.net/client/default

ERDC/CERL TR-20-1

January 2020

Guidelines for Minimizing Potable Water Loss Due to Water Distribution System Flushing

Elisabeth M. Jenicek, Noah W. Garfinkle, and Andy Hur U.S. Army Engineer Research and Development Center (ERDC) Construction Engineering Research Laboratory (CERL) 2902 Newmark Drive Champaign, IL 61824

Victoria Heath Department of Civil and Environmental Engineering Grainger College of Engineering University of Illinois at Urbana-Champaign 205 North Matthews Avenue Urbana, IL 61801

Final Technical Report (TR)

Approved for public release; distribution unlimited.

Prepared for Office of the Assistant Secretary of the Army, Installations, Energy and Environment OASA(IE&E) Washington, DC 20310

Under Military Interdepartmental Purchase Request (MIPR) 11097857, “Development of Guidelines for Minimizing Potable Water Loss due to System Flushing.”

ERDC/CERL TR-20-1 ii

Abstract

Executive Order 13834 (White House 2018) mandates the reduction of wa-ter consumption from potable and non-potable water distribution systems to increase readiness and security. Flushing of water distribution systems potentially wastes large quantities of water. This work developed and ad-ministered an installation survey to assess the prevalence of potable water flushing at Army installations, and used the information gathered from that survey to identify available technologies to recommend best manage-ment practices that reduce flushing and/or that reduce water loss due to flushing. Specific recommendations include changes to lessen the need to flush and practices that optimize necessary flushing.

DISCLAIMER: The contents of this report are not to be used for advertising, publication, or promotional purposes. Ci-tation of trade names does not constitute an official endorsement or approval of the use of such commercial products.

All product names and trademarks cited are the property of their respective owners. The findings of this report are not to be construed as an official Department of the Army position unless so designated by other authorized documents.

DESTROY THIS REPORT WHEN NO LONGER NEEDED. DO NOT RETURN IT TO THE ORIGINATOR.

ERDC/CERL TR-20-1 iii

Contents Abstract .......................................................................................................................................................... ii

Figures ............................................................................................................................................................ vi

Preface ...........................................................................................................................................................vii

1 Introduction

1.1 Background

1.2 Objectives

1.3 Approach

1.4 Scope

1.5 Mode of technology transfer

2 Background on Flushing

2.1 Maintaining water quality

2.1.1 Bacteria control

2.1.2 Water age

2.2 Fire protection measures

2.2.1 Fire flow testing

2.2.2 Valve exercising

2.3 Distribution system maintenance

2.3.1 Main installation

2.3.2 Main repair (planned)

2.3.3 Main break (unplanned)

3 Policies on Flushing

3.1 Federal

3.1.1 Water quality

3.1.2 Fire protection

3.2 Department of Defense

3.2.1 Water quality

3.2.2 Fire protection

3.3 Select state guidelines

3.4 Select local guidelines

4 Potable Distribution System Flushing Methods

4.1 Conventional

4.2 Unidirectional

4.3 Automatic flushing systems

5 Models and Tools

6 Survey of Army Installations

6.1 Survey creation

6.2 Survey administration

ERDC/CERL TR-20-1 iv

6.3 Review of findings

6.3.1 Quantity of potable water used for quality and fire flow flushing vs. population and flushing percent

6.3.2 Effect of management and source water on flushing program frequency

6.3.3 Effect of management and source water on flushing duration deciding factor

6.3.4 Effect of management and source water on flushing reason

6.4 Case studies

6.4.1 Fort Hood, TX: Large site, high flushing (outlier), auto-flusher proposal

6.4.2 Fort Leonard Wood, Missouri: Large site, high flushing (outlier)

6.4.3 Fort Bragg, North Carolina: Large site and low flushing

6.4.4 Fort George Meade, Maryland: Use of Neutral Output Elimination System (NO-

DES)

6.4.5 Fort Greely, Alaska: Small size and low flushing

6.4.6 Fort Irwin, California: Arid climate and regular flushing

6.4.7 Fort Buchanan, Puerto Rico: Small site, low flushing, and non-potable supply

6.4.8 Fort Benning, Georgia: Large population, high flushing, and auto-flushers

6.4.9 Fort A.P. Hill, Virginia: Use of Fire Flow Pro

6.4.10 Fort Huachuca, Arizona: Large site and arid climate

6.4.11 Fort Carson, Colorado: Decreased storage capacity

7 Best Management Practices

7.1 Optimizing treatment and distribution system

7.1.1 Right sizing

7.1.2 Storage tank management

7.1.3 Dead end looping

7.1.4 Pipe material and pipe age

7.1.5 Treatment plant improvements

7.2 Optimizing flushing

7.2.1 Ice pigging

7.2.2 Dynamic flushing plan

7.2.3 Water reuse

7.2.4 Fire hydrant testing

7.2.5 FireFlowPro® diffusers

7.2.6 NO-DES Inc. technology

7.2.7 Auto-flushers

7.2.8 Unidirectional flushing

8 Conclusions and Recommendations

8.1 Conclusions

8.2 Recommendations

Bibliography

Acronyms and Abbreviations

Appendix A: Illinois Section American Water Works Association Survey

Appendix B: Original Project Survey

ERDC/CERL TR-20-1 v

Appendix C: Final Project Survey

Appendix D: Explanation of Altered Survey

Appendix E: Summary of Survey Results

Appendix F: Water Audit Classifications

Appendix G: Federal and DoD Policy

Appendix H: FireFlowPro® Product Information

Appendix I: NO-DES, Inc. Product Information

Report Documentation Page (SF 298)

ERDC/CERL TR-20-1 vi

Figures 2-1 Hydrant flushing at Camp Atterbury 2-2 Fort Bragg potable water with brown discoloration 2-3 Correlation between water age (disinfectant residual as free chlorine) and

TTHM, a regulated disinfectant byproduct 2-4 Correlation between water age (disinfectant residual as free chlorine) and

Haloacetic Acids (HAA5) regulated disinfectant byproducts (USEPA 2013) 2-5 Common water system actions to control nitrification from a 2004 survey 2-6 Old wooden water mains at Blue Grass Army Depot 4-1 Automatic flusher at Blue Grass Army Depot 6-1 Quantity of potable water used for quality and fire flow flushing vs.

population and flushing percent 6-2 Quantity of potable water used for quality and fire flow flushing vs.

population and flushing percent excluding outliers 6-3 Flushing frequency divided by resources management type (left) and flushing frequency by source water (right) 6-4 Flushing duration factor divided by resources management type (left) and flushing duration factor by source water (right) 6-5 Flushing reason divided by resources management type (left) and flushing reason by source water (right) 6-6 Fort A.P. Hill monthly flushing in gallons over the course of 3 years.

Reduction attributed to the use of the Fire Flow Pro diffuser system 6-7 Water consumption (pumped) at Fort Huachuca from 1992 to 2017 plotted against Fort population 7-1 A Texas American Waters Bernard Acres Tank TTHM (and other DBP) removal with pressure differences in spray stripping 7-2 NO-DES truck-mounted system in operation 7-3 Two sets of pipe section before and after NO-DES. Top pictures are before images. Bottom pictures are after NO-DES images

ERDC/CERL TR-20-1 vii

Preface

This study was conducted for the Office of the Assistant Secretary of the Army, Installations, Energy and Environment OASA(IE&E), via Military Interdepartmental Purchase Request (MIPR) 11097857, “Development of Guidelines for Minimizing Potable Water Loss due to System Flushing.”

The technical monitor was Dr. Marc Kodack, Office of the Deputy Assis-tant Secretary of the Army for Energy and Sustainability, ODASA(ES). Fi-nal report review was done by Paul Volkman, ODASA(ES).

The work was managed and executed by the Energy Branch, of the Facilities Division, U.S. Army Engineer Research and Development Center, Construc-tion Engineering Research Laboratory (ERDC-CERL). At the time of publi-cation, Giselle Rodriguez was Chief of the Energy Branch and Dr. Christa Woodley was Acting Chief of the Facilities Division. The Deputy Director of ERDC-CERL was Dr. Kirankumar Topudurti and the Director was Dr.

Lance Hansen.

Special appreciation is owed to the installation points of contact for providing data and information that were invaluable to this study and for reviewing this report. These individuals are: Charles Auer, Mark Fincher, and Tom Horn of Fort Benning, Charles Beach of Fort Knox, Joanna Brown of Fort Riley, John Burgess of USAG Miami, Darryl Butler of Fort Bragg, Melanie Chaballe of USAG Schinnen, Craig Collins and Robert Diaz of White Sands Missile Range, Lissete Cortez of Fort Bliss, Diane Covell of Fort Drum, Bill Douglas of BAE Systems for Radford Army Ammunition Plant, Phil Facer of Fort Campbell, Gef Fisher of Fort A. P. Hill and Gary Manville of American Water for Fort A. P. Hill, Lyle Fogg of Joint Base Lewis-McChord, Steven Golub of the U.S. Army Health Command out of Fort Sam Houston, Beth Grosscup of Fort George Meade, Jim Casey and Vincent Guthrie of Fort Carson, Christian Hawkinson of Rock Island Arse-nal, Paul Herzer of Carlisle Barracks, Christopher Higgins of Fort Huachuca, Luke Horace of the City of Augusta Utility and Glenn Hartzog for Fort Gordon, Linda Howard of Fort Polk, Beckie Hurley of Hurco Tech-nologies, James Johnson of USAG Daegu, Mark Lewis of Fort Detrick, Mike Maier and David Ondre of Picatinny Arsenal, Dennis Meyers of Fort Sill, Mike Miller of Fort McCoy, Anibal Negron of Fort Buchanan, Daniel O'Brien of Fort Belvoir, Gary Ogden of Fort Lee, Sang Hyon Pak of USAG Yongsan, Keith Pendleton of Fort Leonard Wood, Tami Relph of the Naval Facilities Engineering Command at Port Hueneme, Chad Reynolds of

ERDC/CERL TR-20-1 viii

American Water for Fort Hood, Fred Sandgren of Fort Wainwright, Steven Sheffey of Dugway Proving Grounds, Gregory Skaggs of Yuma Proving Ground, Norbert Soehn of USAG Heidelberg (I), Al Thomas of Fort Greely (Water Contractor), Richard Valcourt of the Natick Soldier System Center, President Chris Wilkinson of NO-DES Inc., Randall Wilson of Redstone Arsenal, Brett Wobker of Fort Leavenworth, Chris Woodruff of Fort Irwin, and Giovanni Zanon of USAG Vicenza.

COL Teresa A. Schlosser was Commander of ERDC, and Dr. David W.

Pittman was the Director.

ERDC/CERL TR-20-1 1

1 Introduction

1.1 Background

The flushing of water distribution systems on Army installations is an es-sential part of a process that maintains the systems in optimal condition and that ensures installations’ fire protection capabilities. Army Regula-tion (AR) 420-1, Army Facilities Management (HQDA 2008) requires flushing. However, unsophisticated flushing protocols can potentially cause significant water losses that can affect an installation’s total water usage (PNNL 2012a-f, PNNL 2013a,b). There is a need for best manage-ment practices (BMPs) that can significantly reduce water loss while main-taining regulatory compliance. This work was undertaken to review and optimize the flushing of distribution systems to help installations achieve and maintain system stability and readiness.

1.2 Objectives

The overall goal of this work was to inform resource management to main-tain installation readiness through the development, protection, and resto-ration of the nation’s water (U.S. Army War College 2015). Specific objec-tives were to evaluate and recommend BMPs that will reduce the loss of potable water during water distribution system flushing.

1.3 Approach

The objectives were completed through the following steps:

1. A literature review was conducted to explore flushing technologies available to minimize water loss due to flushing.

2. A review was done of relevant federal, state, local, industry, and Army policies.

3. An installation survey was developed and administered to assess the prevalence of potable water flushing at Army installations (see section 6.4, “Case studies”) and to identify available technologies that reduce flushing and/or that reduce water loss due to flushing.

4. BMPs were identified and documented.

ERDC/CERL TR-20-1 2

1.4 Scope

This project focused on the outcomes of past water use studies and data disaggregation efforts identified by the sponsoring organization’s repre-sentative. This work could be replicated elsewhere by considering unique aspects of each installation’s water system.

1.5 Mode of technology transfer

It is anticipated that the information gathered and generated by this work will support changes to policy and specifications, and will inform efforts to improve water resilience, to enhance water security (and the allocation of security funds), and to reduce water consumption.

ERDC/CERL TR-20-1 3

2 Background on Flushing

Water distribution systems have been developed and operated for a variety of reasons, one of which is to distribute potable water. In some places, the need for non-potable water is great enough to support alternative water supply configurations, most commonly by operating dual (potable and non-potable) distribution systems. However, in most locations water sys-tems established to provide potable water for household use are aug-mented and enlarged to provide water for irrigation, car washing, indus-trial processes, fire protection/suppression, and other purposes. This re-port focuses on flushing of potable water distribution systems.

Since multiple uses rely on the same distribution system, those users must work together. Notably, the water utility has important partners those that provide fire protection. The water utility must work together with local fire officials, building code officials, and others to effectively and efficiently promote fire protection. Good communication between all involved parties is essential to protect property and life.

When a distribution system is flushed, it loses a variable fraction of the total potable water input to the system. It can be difficult to obtain an exact meas-ure of water consumption and losses at installations since individual facilities are metered.* However, a review of water consumption at some of the eight pilot net zero water installations (PNNL 2012a-g, PNNL 2013a-b) revealed that the average percentage of the end-use “loss” and “unknown” categories (of which flushing is a part) totaled 24.6% of total water consumption. This amount corresponds with results of discussions at the FY2012 Net Zero In-stallations Workshop where installation water managers named water distri-bution system condition/water loss as their biggest concern.

Similarly, water loss was shown to be a significant component in the in-stallation water models that the Engineer Research and Development Cen-ter, Construction Engineering Research Laboratory (ERDC-CERL) created for 29 Installation Water Sustainability Assessments under the sponsor-ship of the Office of the Deputy Assistant Secretary of the Army for Energy and Sustainability (ODASA[ES]). At some of these sites, distribution sys-tem flushing alone accounted for up to 30% of the total installation water

* AWWA’s M36 Water Audits and Loss Control Programs contains the accepted methods for quantifying authorized consumption, unauthorized consumption, and apparent losses, which are used to calculate real losses (leakage).

ERDC/CERL TR-20-1 4

use. These large measured losses justify an investigation into why flushing occurs and how the practice can be best managed to reduce losses.

Flushing of water systems on Army installations takes place for a variety of reasons. Certain degrees of distribution system flushing are mandated by federal law, state law, local jurisdictions, and Army policy (Policies on Flushing). Flushing is typically required to protect water quality, to ensure fire flow (Figure 2-1), and to remove contaminants introduced during con-struction or system maintenance. The following sections summarize com-mon reasons for system flushing. Note that the listed reasons are not ex-haustive; individual systems often have unusual and unpredictable needs for flushing.

Figure 2-1. Hydrant flushing at Camp Atterbury.

2.1 Maintaining water quality

Many water utilities operate flushing programs to maintain good water quality in the distribution system (AWWA 2016). Water quality within the distribution system is monitored by testing for bacteria, which ensures control of current bacteria levels, and by testing for disinfectant byprod-ucts and chlorine residuals, which ensures the control of future bacteria growth as a function of water age.

2.1.1 Bacteria control

The purpose of a potable water distribution system is to supply safe drink-ing water for consumption. However, bacteria-harboring films accumulate

ERDC/CERL TR-20-1 5

in the system piping over time. The presence of bacteria and other patho-gens can seriously jeopardize water quality. For this reason, bacteria levels are highly regulated; the control of bacteria and prevention of other patho-gens is crucial to supplying potable water.

Flushing is commonly used to control bacteria formation since flushing scours away the bacteria-harboring films and replaces any water in the lines with fresh finished water. Since the bacteria-harboring films usually re-develop over time, flushing is not a permanent solution to the problem of bacterial contamination. The process must be performed at a frequency that matches the film’s development.

2.1.1.1 Sedimentation

Sediments can accumulate in the pipes due to corrosion and scaling (AWWA 2017b). Other unusual circumstances can also cause sediments to infiltrate the system. For example, nearby construction or vibrations in the ground can shake loose otherwise bound deposits.* Removal of solids and scouring of water mains keeps pipes in working order by maintaining a clear flow path and removing bacteria-hosting deposits. To properly con-duct flushing operations for this purpose, the water flowing in the pipes must reach velocities capable of pushing the otherwise settled sediments out of the lines. These higher-than-normal velocities scour the insides of the distribution piping, helping to clean out any accumulated soils or silt.

2.1.1.2 Tuberculation

Tuberculation is the formation of small knob-like rust formations (tuber-cles) usually in unlined cast iron piping (AWWA 2017b). Flushing at higher-than-normal water velocities can scour the inside of water lines and thereby remove iron deposits that would have otherwise harbored bacteria or led to further corrosion by accumulation. The presence of tubercles can be attributed to the pipe material used, and to the pH or composition of the source water, or to the presence of stagnant water sections.

* Personal communication between Victoria Heath, ERDC-CERL and Beth Grosscup, Fort George Meade, Maryland, 2018.

ERDC/CERL TR-20-1 6

2.1.1.3 Biofilm

Bacteria are also prevalent in biofilms that adhere to the inner walls of dis-tribution mains. The accumulated film can host a number of bacteria; re-moval of the film can help control bacteria in the system. The biofilm can host coliforms and jeopardize the total coliform standard. The prevalence of biofilms increases in conditions of low flow and high-water tempera-tures, and in the presence of pipe corrosion (Ohio AWWA 2014). Biofilms are most commonly removed by flushing at a velocity that scours the inner walls of the pipe. Their presence can be attributed in part to the pH or composition of the source water, and to the amount of time stagnant water stands in the pipes.

2.1.1.4 Customer complaint

Customer complaints are usually based on water taste, color (Figure 2-2), or odor. Most of these conditions can be attributed to the films just de-scribed (sedimentation, tuberculation, or biofilm). By keeping good rec-ords of complaints, installations can determine concentrated areas of the complaints. Focusing attention on these problem areas can help identify specific problems. For example, water quality problems commonly occur on the system’s periphery where stagnant water stands and films accumu-late. Flushing removes many of the deposits, sediment, and other materi-als that can affect taste and odor. In many complaint-related cases, spot-flushing or conventional flushing is used to flush out the foul water and draw in fresh. This does not conserve water, but it does preventively ad-dress the issue. The only other way to improve taste and odor would be to prevent stagnation and the accumulation of films.

2.1.2 Water age

Water’s quality diminishes with its age. As water sits unused in pipes, it ages and stagnates. Finished water in the distribution system can stagnate for a number of reasons (low demand, oversized pipe diameter, dead-end loca-tion, loss of pressure, etc.). Stagnant water can increase the production of bacteria-harboring films and the conditions that cause stagnant water also allow time for undesirable reactions between the water and its disinfectant.

ERDC/CERL TR-20-1 7

Figure 2-2. Fort Bragg potable water with brown discoloration.

Operational events can also contribute to the conditions that cause stag-nant water, e.g., when buildings remain unoccupied for long time periods, a common occurrence at force projection platforms. The ability for bacte-ria to form in this stagnant environment is only limited by the presence of chlorine disinfectant (added during the treatment process). As the disin-fectant residual decreases (and is used) bacteria can begin to grow.

There are only two ways to truly achieve a net reduction in water age in a system, by increasing demand or by reducing the storage volume of the sys-tem (AWWA 2017a). Flushing can be conducted to remove old water, essen-tially creating an artificial demand (that will reduce stagnation). The condi-tions of the aged water can also be addressed by treating its specific undesir-able qualities (including disinfection byproducts and High Nitrates).

2.1.2.1 Disinfection byproducts

Water treatment disinfection can generate contaminants called disinfection byproducts (DBPs) that are caused by the reaction of source water with treatment chemicals (Hruday and Charrois 2012). These reactions occur as water ages and their concentration can be graphed over time (Figures 2-3 and 2-4). Note that Figures 2-3 and 2-4 graph the correlation between wa-ter age and only two regulated DBPs (Trihalomethanes [TTHM] and volatile HAA5s). (Many more individual species exist.) Disinfection byproducts can be controlled by flushing the system of aged water, by making changes to the treatment regimen, and by making engineered changes to the distribu-tion system. Treatment process improvements include the use of alternative disinfectants and pre-oxidants, aeration, and or organic matter removal

ERDC/CERL TR-20-1 8

(enhanced coagulation, activated carbon, ion exchange, high-pressure membranes). Aeration helps to remove volatile DBPs only. Chapter 7, “Best Management Practices,” discusses these removal techniques in more detail.

Figure 2-3. Correlation between water age (disinfectant residual as free chlorine) and TTHM, a regulated disinfectant byproduct.

Source: AWWA 2016b.

Figure 2-4. Correlation between water age (disinfectant residual as free chlorine) and Haloacetic Acids (HAA5) regulated disinfectant byproducts (USEPA 2013).

Source: AWWA 2016b.

ERDC/CERL TR-20-1 9

2.1.2.1 High nitrates

Nitrate formation in the distribution system is essentially a function of wa-ter age and stagnation in systems that use chloramine disinfectant (AWWA 2017b). When nitrification episodes occur, similar to disinfectant byprod-ucts, the water in the lines should be removed and replaced with fresh fin-ished water. While flushing is considered the most effective response to ni-trification, other water system actions are available (Figure 2-5).

2.2 Fire protection measures

Although there are exceptions, the potable drinking water distribution sys-tem and the fire protection water distribution are commonly the same sys-tem. Assuming the two systems are the same, the same lines would be re-sponsible for maintaining legal potable quality and for providing the appro-priate quantity of water needed for firefighting. Firefighting capability is measured/maintained in two ways: (1) by ensuring that all parts of the sys-tem open and close as expected (via valve testing), and (2) by leaving the hy-drants open long enough to ensure that a pre-determined amount of water and pressure will be delivered (fire flow testing). Note that fire hydrant flushing should include flush-hydrants, which are named for their location below ground (level or “flush” with the surface). Flush-hydrants, which are usually found in airfields, should be regularly tested (AWWA 2010).

Figure 2-5. Common water system actions to control nitrification from a 2004 survey.

Source: Kirmeyer 2004

ERDC/CERL TR-20-1 10

2.2.1 Fire flow testing

Fire flow tests determine the pressure and flow-producing capacities at any point in the distribution system. Fire flow testing should be conducted every 5-10 years per location (AWWA 1989) or every 5 years as recom-mended by the National Fire Protection Association (NFPA) code 291 (NFPA 2016). Fire flow testing, also called main capacity flow testing, checks for the hydrants’ water production ability (AWWA 1989). Hydrants are opened and tested to see if they meet their full fire flow, i.e., their abil-ity to sustain a certain pressure for a given time period. These pressure tests confirm that the hydrant can dispense the volume of water needed to fight a fire of the size determined to be possible in that location. This can also be related to reserve waters needed for fire protection. Sometimes a storage tank is necessary to ensure that enough water is available in the event of a firefighting emergency. This flushing purpose is difficult to re-place by other means; essentially flushing is the only viable option for compliance. Hydrant flow testing results can also be used in system plan-ning for model calibration.

2.2.2 Valve exercising

Each distribution system hydrant and valve should be inspected annually to ensure operation. Locations in areas of freezing temperatures may want to conduct this twice per year (AWWA 1989). This inspection simply opens the valves or hydrants for a moment to prove them operational. Operating fire hydrants and valves helps to ensure that they will work properly in an emergency. In some cases, while the hydrant is open anyway, water is al-lowed to run until it is clear, in a combination of quality-related flushing and valve exercising. Flushing is currently the only means to truly exercise valves/hydrants. Fortunately, the quantity expelled is low relative to other flushing purposes since valve exercising only tests operation, not pressure or quantity delivered.

2.3 Distribution system maintenance.

Flushing is a required part of infrastructure maintenance. Essentially, pipe that has been recently worked on must be flushed to remove any contami-nants introduced during construction. The aging infrastructure across the country (e.g., see Figure 2-6) makes this a potentially major flushing compo-nent. Any loss of pressure in the system will also trigger a flushing require-ment and will result in a ‘boil order’ for any human consumption of the water.

ERDC/CERL TR-20-1 11

Figure 2-6. Old wooden water mains at Blue Grass Army Depot.

2.3.1 Main installation

ANSI/AWWA Standards 600-606 (AWWA 2002) address pipe installation and recommended flushing procedures. This necessary flushing is done to sanitize new materials in contact with potable water even if precautions have been taken to reduce contamination (such as sealing the ends of new sections and/or the use of high chlorine concentrated preparatory spray).

2.3.2 Main repair (planned)

Another potential source of contamination is through breaches in the sys-tem caused during repairs. The amount of repair work required to main-tain the aging infrastructure across the United States and within U.S.

Army installations is such that, at some locations, water loss from flushing following repairs can be great.

Such repairs can come in the form of pipe length replacement, or of changes from one type of pipe material to another. So, for example, it might make more sense in some repair scenarios to line a main rather than to replace it completely. This is a feasible solution to replace easily cor-roded ductile-iron mains, which can be lined with a cement layer that pro-tects against corrosion and would theoretically reduce the frequency of flushing required for the removal of corroded particulates and sediments (ANSI/AWWA n.d.). In general, these replacements require the same amount of flushing (AWWA 2008b).

ERDC/CERL TR-20-1 12

2.3.3 Main break (unplanned)

In general, it is easier to disinfect a new main installation through repeat flushing than it is to do so with a break. The unsanitary conditions created when an existing main bursts is difficult to control (ANSI/AWWA 1993).

Unfortunately, breaks are extremely common and are estimated to occur once per year per every 4 miles of pipe in the average distribution system (AWWA 2002). The leading cause of such breaks is improper installation.

Breaches in the system, whether from scheduled repairs or unplanned pipe ruptures, must be followed with flushing.

ERDC/CERL TR-20-1 13

3 Policies on Flushing

Flushing is a tool commonly used in water distribution systems on Army installations to protect water quality and to ensure installations fire pro-tection capabilities. To meet those goals, certain policies and regulations (at the federal, state, local, and Army level) explicitly require flushing on a routine basis. Appendix F includes a listing of federal and U.S. Department of Defense (DoD) policies. Note that the listing does not include state and local policies since this work reviewed only select regulations; many more exist at the country and global scale.

3.1 Federal

3.1.1 Water quality

The primary regulation governing water quality is the 1974 U.S. Environ-mental Protection Agency (USEPA) Safe Drinking Water Act (SDWA 1974). The SDWA regulates health-based standards for drinking water to protect against man-made and naturally occurring contaminants. Regu-lated impurities include chemicals, animal waste, pesticide, human threats, wastes injected underground, and naturally occurring substances.

The SDWA sets national enforceable contaminant levels and/or required ways to treat water to remove contaminants. Contaminants are divided into a group of primary regulations and a set of secondary drinking water standards; the second of which is non-enforceable. These secondary, non-life-threatening contaminants include clarity, bad taste, and odor, all of which are regularly addressed by flushing.

The SDWA requires the collection of drinking water data and the monitor-ing of contaminants. The act also oversees state drinking water programs.

The SDWA applies to public water systems that supply Army installations and to Army water systems through call-out in regulation and Unified Fa-cilities Criteria (UFC). Monitoring or compliance is essentially maintained through the measurement bacterial counts.

Several crucial amendments have been made to the SDWA since its incep-tion. Before 1990, water quality standards were applied at the treatment plant. In 1990, the SDWA was amended to require that water quality also be monitored at the most distant outlet of the water distribution system, ERDC/CERL TR-20-1 14 ensuring that water quality standards apply from the water source to the tap. This amendment requires that adequate disinfection exist at the most remote point of withdrawal.

In 1996, the SDWA was amended to require that consumer confidence re-ports of water quality be composed and distributed to the public served.

The reports provide guidance, assistance, and public information about drinking water. Water systems are also required to notify customers if any standard is not met.

In 2002, the SDWA was again amended to counter bioterrorism threats.

Public Law 107-188, Public Health Security and Bioterrorism Prepared-ness and Response Act amended the SDWA to require Water System Vul-nerability Assessments (WSVA) and Emergency Response Plans (ERPs).

Another set of federal regulations governing water quality are the Stage 1 DBP Rule (40 CFR 141-63, FR 69390, 1998) and the Stage 2 DBP Rule (40 CFR 141-71, FR 388, 2006). These rules aim to reduce exposure to DBPs by setting enforceable standards and by requiring compliance through regular monitoring. These measures of compliance include a limit on DBP formation a more stringent minimum disinfectant residual re-quirement.

3.1.2 Fire protection

There are several federal fire protection related policies. NFPA code 291, Recommended Practice for Fire Flow Testing (NFPA 2016), provides guidance on fire flow tests and marking of hydrants to determine available fire service water supply from hydrants. The code recommends that stand-ard flushing practices for fire flow testing be conducted every 5 years to verify capacity, and that valve exercising be done annually to identify defi-ciencies that could be corrected to ensure adequate fire flows.

The American Water Works Association (AWWA) standards for fire protec-tion include AWWA M17, Installation, Field Testing and Maintenance of Fire Hydrants (AWWA 1989), and AWWA M31, Distribution System Re-quirements for Fire Protection (AWWA 2008a). The AWWA has published standards for materials used in the field of fire protection for many decades.

However, it was not until the early 1980s that AWWA’s Committee on Fire Protection developed the first edition of this manual, which addresses the

ERDC/CERL TR-20-1 15

planning, design, and maintenance of distribution systems that supply wa-ter for fire protection and suppression. This manual provides guidance on the design, operation, and maintenance of water distribution systems as they relate to fire protection and fire suppression activities. When the gov-erning body of a community makes a conscious decision to use the available water supply system for fire suppression purposes, this manual should be consulted, judiciously applied, and tempered as local conditions require.

The manual is intended to help water utilities in the design and operation of their systems to maximize fire protection benefits while delivering safe, po-table water to customers. Note that this manual emphasizes public water systems rather than water systems exclusively designed for fire protection.

Also, users should note that this manual does not reproduce material availa-ble elsewhere in AWWA publications such as Manual M17, Installation, Field Testing, and Maintenance of Fire Hydrants (AWWA 1989), Stand-ards C502, Dry-Barrel Fire Hydrants (AWWA 2014a), and Standards C503, Wet-Barrel Fire Hydrants (AWWA 2006).

The adequacy of a water distribution system for fire protection depends on the fire flows required. Chapter 1 of AWWA M31, Distribution System Re-quirements for Fire Protection (AWWA 2008a) describes several methods for determining required fire flows. Once fire flow requirements are deter-mined, these are added to the other water system demand requirements.

Chapter 2 discusses the impact of fire protection on distribution system de-sign. Chapter 3 focuses on distribution storage in terms of both sizing and lo-cation. Chapter 4 discusses reliability issues arising because systems must re-main in operation even when individual components are out of service. Chap-ter 5 describes fire suppression sprinkler systems and their effect on water re-quirements, particularly because they affect how water is used for firefighting.

Appendix A of AWWA M17, Installation, Field Testing and Maintenance of Fire Hydrants (AWWA 1989) lists organizations involved in fire protection and describes the organizations and their roles. Appendix B describes the relationship between the water supply system and fire insurance ratings, because one benefit of supplying needed fire flow is an improved fire in-surance rating for a community.

ERDC/CERL TR-20-1 16

3.2 Department of Defense

3.2.1 Water quality

Army Regulation 420-1 (HQDA 2008) provides policies governing the op-eration, maintenance, and repair of water supply systems. Installations are required to prepare a Water Resource Management Plan to include an ef-fective water conservation program. Violations of the SDWA will be re-ported using the Army Compliance Tracking System. This regulation also requires new and repaired water mains, storage tanks, wells, and equip-ment to be disinfected in accordance with AWWA standards C651

(ANSI/AWWA 1993), C652-92 (ANSI/AWWA 1992), C653-13 (AWWA

2013), and C654-13 (ANSI/AWWA 2013), following construction, repairs, installation of taps, or contamination situations.

UFC 3-230-01, Water Storage, Distribution, and Transmission (USACE et al. 2014) offers recommended design guidelines for the distribution sys-tem. Design often determines the frequency and location of flushing pro-grams attempting to eliminate stagnant water.

UFC 3-230-02, Operation and Maintenance: Water Supply Systems (USACE et al. 2001) recommends that flushing for sediment removal be done annually in a unidirectional way or in a way that provides adequate scouring velocities. The UFC also suggests that dead ends be flushed, mini-mally, once a year, and that flushing in response to odor or taste com-plaints be done as needed.

3.2.2 Fire protection

UFC 3-600-01, Fire Protection Engineering, Standard (USACE et al.

2013) applies to planning and development projects and to the design con-struction and commissioning documents used for the procurement of fa-cilities. Primary fire protection criteria reference design-bid-build and de-sign-build contracts. This UFC specifies rules related to hydrant capacity for a given number and size of structures, which determine the capacity of the distribution system. The UFC also includes a requirement that fire de-partments perform annual fire flow testing (U.S. Army 2009).

UFC 3-601-02, Operation and Maintenance: Inspection, Testing, and Maintenance of Fire Protection Systems pertains to the inspection, testing, and maintenance of engineered fire protection features in DoD facilities.

The UFC includes the mandate for WSVAs and ERPs, and lists flushing as a common task for use in inspection, testing, and maintenance.

ERDC/CERL TR-20-1 17

3.3 Select state guidelines

State guidelines related to flushing vary across the country. The following paragraphs include examples of state laws that would require installation compliance. Many state guidelines reference (or simply echo) the require-ments of federal law. Other state laws adopt more restrictive regulations based on the state’s specific goals (usually originating from state’s depart-ment of public health or the state’s interpretations of the SDWA).

California, for example, has state-specific flushing requirements in accord-ance with the California Safe Drinking Water Plan. One California utility, the Western Municipal Water District (WMWD 2016a), informed its resi-dents that their water system flushing program helps to maintain high qual-ity water throughout their service area and that flushing operations are re-quired on a regular basis by the California Department of Public Health (WMWD 2016b).

In the state of Florida, public water suppliers that claim to provide fire protection are required to have finished water storage tanks or reservoirs with sufficient capacity for fire prevention (FRWA 2012) even though they are not required to make any such claim. This regulation is important be-cause it attempts to control the amount of storage, which is known to have a great impact on flushing requirements if supply exceeds demand.

3.4 Select local guidelines

In some locations, even more location-specific regulations related to flushing exist beyond federal, state, and Army policy. Installations should be aware of and consider county-level regulations (within the United States), or host countries regulations (outside the United States) when optimizing flushing.

An example of country specific law can be seen in Italy where water rights and infrastructure ownership are highly contested issues (Caporale 2015).

Such issues can complicate distribution system maintenance and subsequent flushing programs by adding a layer of compliance. In Italy, U.S. installations must be sensitive to local requirements and thoroughly investigate local laws when planning, developing, and implementing a flushing program.

Some contracts between Army installations and the local governments or private utility contractors may require flushing protocols not specified

ERDC/CERL TR-20-1 18

elsewhere in criteria or regulation. For example, Fort A.P. Hill conducts fire flow testing annually opposed to the standard of about every 5 years.*

Additionally, area-specific water withdrawal permits may be in place in some locations. This limitation on withdrawals has a great influence on the operation and maintenance of the distribution system and essentially holds the utility more accountable for water-conserving usage. However, it is pos-sible for the installation to be exempted from the permitting law, which in turn would provide little incentive for conservation. This is exemplified in a part of the state of Virginia where groundwater withdrawal permits are re-quired for all facilities east of interstate 95. Even though Fort A.P. Hill is lo-cated east of the highway, the installation is exempted from permitting.†

* Personal communication between Victoria Heath, ERDC-CERL and Gary Manville. Fort A.P. Hill, Virginia, 2018.

† Personal communication between Victoria Heath, ERDC-CERL and Gary Manville. Fort A.P. Hill, Virginia, ERDC/CERL TR-20-1 19

4 Potable Distribution System Flushing Methods

Flushing is recommended at several stages of the distribution system’s life

• initially after installation (per AWWA C600-93 [AWWA 1993])

• as needed after install for scouring and water quality control (per

AWWA C652-92 [ANSI/AWWA 1992] and AWWA C651-92

[ANSI/AWWA 1993])

• if repair or maintenance to the pipe is ever needed (Environmental Protection Agency)

• annually for hydrant inspection

• every 5-10 years for hydrant flow testing (AWWA 1989, NFPA 2016).

The many possible purposes for flushing can make it a challenge to de-velop a comprehensive program. However, a comprehensive plan is key to an efficient and water-conserving flushing operation. This objective (to achieve efficient and water-conserving flushing operation) should drive the plan creation (Martin and Ries 2014). The following sections describe common approaches to flushing programs.

4.1 Conventional

Conventional flushing, sometimes called reactionary hydrant flushing (Reiss et al. 2010), involves opening just a few specific hydrants, usually in no determined order for an indeterminate amount of time. Flushing of this type is typically conducted in response to a customer complaint or on a loose annual basis. In either case, hydrants are typically opened and left to run until the water appears clear and color visibly improves. The velocities achieved in conventional flushing are not usually sufficient to remove loose pipe deposits (Reiss et al. 2010). Conventional flushing promotes water re-placement but does not generally result in scouring (Mercer 2010). In con-ventional flushing, no valves are closed to increase velocity (Martin and Ries 2014). Conventional flushing requires as much as 40% more water than does unidirectional flushing (Boulos et al. 2006). As water conservation has been prioritized over the last few decades, localized conventional flushing is used more commonly as a response to quality test failures or customer com-plaints than as a method for annual flushing. The use of conventional flush-ing is not favored as part of a routine or annual program on the system wide scale because the practice has many disadvantages.

ERDC/CERL TR-20-1 20

4.2 Unidirectional

Unidirectional flushing (UDF) is a systematic approach to flushing that in-volves specific design sequences (Reiss et al. 2010). In this case, the entire distribution system or sub network (loop division) is flushed from the clean water source out to periphery of the distribution network. This directional flow keeps debris that is being removed from re-entering a clean stretch of system. UDF achieves higher velocities, between 5-10 feet per second (Reiss et al. 2010), enough to scour and remove biofilms and/or sediments.

This type of systematic approach requires the development of a plan or pro-gram that will likely need to be revised every year or so. The program is de-veloped from a hydraulic modeling of the distribution system. An engineer can use a current model of the distribution system assets to isolate pipe seg-ments of interest by valve closing to achieve the desired pipe velocities (Reiss et al. 2010). This often requires more upfront time (Mercer 2010).

In UDF, after the valve sequence has been established, flushing time is usually based on the removal of two to three volumes of water. However, visual inspection usually takes final precedence. If discoloration persists after the pre-determined length of time for volume removal, field crews may need to extend flushing times. UDF success in a city case study showed an 85% reduction in iron, 32% increase in chlorine residual, 75% drop in turbidity, and 78% reduction in complaints after UDF was per-formed, relative to conventional flushing alone (Reiss et al. 2010).

4.3 Automatic flushing systems

Automatic Flushing Systems (AFSs, see Figure 4-1) can be used when ve-locities (to achieve a scouring effect) are not of concern and when the flushing purpose is simply for water turnover. AFSs can be used in two ways: as continuous blow-off points that are essentially always flushing, or as periodic flushes programed on a daily/weekly/as-needed basis. This is considered a short term (preventive) strategy (Boulos et al. 2006). Auto-matic flushing can optimize flushing water consumption because regular periodic flushes can be smaller than manual flushes (Mercer 2010). Auto-matic flushers are often employed at dead ends in the system or in areas of low demand. AFSs essentially take the place of looping the infrastructure or creating an artificial demand.

ERDC/CERL TR-20-1 21

Figure 4-1. Automatic flusher at Blue Grass Army Depot.

ERDC/CERL TR-20-1 22

5 Models and Tools

Models and tools for water quantity and quality control are not only be-coming more popular but more of a necessity for the sophisticated water provider. Models vary in their capabilities, but ultimately provide a greater understanding of the distribution system; some can even simulate scenar-ios for more proactive water management. Note that pressure and velocity within a pipe are inversely related; as one increases, the other decreases (AWWA 2016). Software that allows the user to visualize and conceptual-ize the system can make these types of relationships more manageable.

The hydraulic integrity of a system depends on its ability to provide a relia-ble supply that meets all demands by providing adequate pressure, fire pro-tection, and uninterrupted reliability (Boulos et al. 2006). Hydraulic models are used to design distributions systems and to modify and maintain them after design. A well-modeled system can form the basis for a highly sophisti-cated flushing plan that optimizes flushing efficiencies. Though models can be costly, a well-chosen mode can ultimately provide an invaluable source of information and a dynamic tool that gives management more sophisticated system operation. These engineering-based models incorporate both physi-cal and flow related characteristics of the system. Hydraulics can be mod-eled using Computational Fluid Dynamic (CFD) modeling.

Water quality can often be inferred from the hydraulic model. However, while the physical hydraulics of the system can be computerized and ma-nipulated to optimize pressure and quantity, a more powerful tool would be one that incorporates the results of water quality monitoring as a means of forecasting quality issues. Knowledge of water quality parameters in ref-erence to space and time can be used to create a tool that can help with op-erational management of such parameters as water age control, nitrifica-tion, mixing issues, booster chlorination, and more. This type of program could help create a maintenance schedule for flushing, tank cleaning, main repair pipeline rehabilitation, and more.

Models already in use by installations include but are not limited to:

AquaGEO (formerly Aqua CAD Suite), EPANet, InfoWater, Water CAD, and Water GEMS.

ERDC/CERL TR-20-1 23

6 Survey of Army Installations

6.1 Survey creation

A questionnaire was created, based on an Illinois Section of the AWWA (ISAWWA) Flushing Survey (Appendix A). While the questionnaire was underway, the need for important changes to the survey became clear. Ap-pendix B contains the original project survey. Appendix C contains the fi-nal version of the project survey that was used for data analysis. Appendix D lists the changes made and explains the need for those changes. Appen-dix E lists the compiled survey results.

6.2 Survey administration

The survey was administered to 156 Army installations (Forts, Arsenals,…